PyPI · #1068 most downloaded on PyPI

faiss-cpu

1.15.112M downloads/mofacebookresearch/faiss ↗New releases in your daily digest, changelog included.

A library for efficient similarity search and clustering of dense vectors.

Last release 18 days ago
16 Sep 2026
Ships fairly regularly
a new release about every 2 months
Most releases are documented
notes for 20 of 30 stable releases
Nothing withdrawn
no release was ever pulled
7 years old
30 releases · first in 2019

Release timeline

30 releases since 2019

20192020202120222023202420252026

Releases

1.15.116 Sept 2026
See CHANGELOG.md for the full list of changes.
See CHANGELOG.md for the full list of changes.
Open source →
Additional notes2 sources agree
Added

3b3711c112fbf7dfb9c64b6fcf90ffc3ca412548 Add explicit IVFPQ scanner distance modes (#5632)

4fdbc9dc18adbfef3cd410a2446aa4c757320a8a Support ZeroCopyIOReader for IndexRaBitQ deserialization (#5629)

b5a14632a6f2bfdc824c86d683b9d6e9205c2604 Run the eight SIMD kernel tests that no build executed (#5620)

e6bd1d252caa5b9315a04390269d382845984ae2 Add AVX512 VPOPCNT dynamic dispatch level (#5531)

7f4cedda9c384de652c7bd5690db0f808d668216 Add staged RaBitQ storage to HNSW (#5526)

cfbf4ec5ff68d0bca4701d0b24b8a0736e110932 SVS: allow option to disable vector storing (#5525)

cda4f6e7133c629e502c2ea9bc51e0ef51847ab1 Expose IndexBinaryIVF.set_direct_map_type via the C API (#5451)

256980cb6ea6610e4b0edf497a038086df8bbfce Add SVE low-dimensional nearest fast path and spherical SuperKMeans support (#5530)

3a716ead1a3c4562fe6e96d982d26789c58e19ed Add Windows ARM64 support to CI (#5544)

7059eaf7da7eddda62e71367e684d4bdedd7f94f Add AVX512 bitplane kernel for multi-bit RaBitQ inner product (#5522)

a424dcb809fd725c44dd976d9063febd4837d16a faiss HNSW: add opt-in deterministic lock-free graph build (faiss::hnsw_deterministic_build) (#5486)

da3191e754b49ca67b6503a9262e96ce5a3d8ed8 IDSelectorWithContext: extend context hook to IVFPQ/PQR, multibit RaBitQ, and Panorama scanners (#5491)

edddbc998e858f16330f9b226dfd72f73d4d1d46 IDSelectorWithContext: scan-context hook for locality-aware selectors (#5490)

Changed

3bb48643ab8ea9e50162b354d2463270acd6ed9c Pin faiss-gpu-cuvs RAPIDS deps to the 26.06 build series (#5643)

6bbb068ad8dab612b84304400846da1b14539085 Measure 20-byte Hamming codes eight at a time (#5587)

3c7042f17a467c38d99f490ef39e4dfa615dc510 Let the binary Hamming paths reach the VPOPCNT kernels

9f993087b07e83f026248addb01dd99e99b15a3e Name the SIMD level masks after what they hold (#5600)

e2cf9e1ebf24c14b888fa69eee96713aa6d64b1a Accept ragged code sizes in faiss::hammings(), and fix the vectorization axis for wide codes (#5520)

545550f6a27031c2269881d97237fd6a5bef15df Use unittest assertions and f-strings in test_index_composite (#5571)

af9eeffcf70d7795107c2b0511bdba507ac45339 Reserve cs vector capacity in hammings_knn_mc_impl to avoid reallocations (#5548)

317a207bcacfcfa695d98289c3f4ea562b782837 ARM SVE optimization for L2 distance kernels adapted to new simd_impl structure (#5098)

2027d3b33144dbf19674b540ec0e31b2f842b1cc Make cuVS dispatch opt-in rather than build-implied (#5529)

a39e2606887bbacc6c8f54e783c4b42da00cf4b3 Fold multi-GPU CAGRA build into train(), delete trainMultiGpu (#5521)

1f93154314afbef210f0ebebeab840da22f9ec7d Revert D114685755: Fold multi-GPU CAGRA build into train(), delete trainMultiGpu

6644dcfa5576221b1a4417fd766a8b2af71d869e Fold multi-GPU CAGRA build into train(), delete trainMultiGpu (#5500)

80a16564f86530dbf0bfaf96c2b71feffeb5093f Enable Faiss dynamic dispatch for OVIS and VeST fbpkgs (#5515)

2688c34f942c8462d97c8632c720a53653456270 Parallelize the k-means++ D^2 seeding update (#5457)

76c67b7b736ca80c2b3ea29528434d6583adc88b Update default Panorama batch size to 1024 (#5441)

3b50babc5914570576b09da043f64dfd7d62a075 Optimize RVV batch-4 distance kernels (#5469)

Fixed

7c3b55c01b8418dc044881625a260fceae9c712b Make Panorama bounds checks overflow-safe (#5616)

fb2d57a3121c073828d2ed7d04ccf09d8bea771c Fix leak and unbounded lists[] in OnDisk invlist deserialization (#5560)

fc13c81fb8b22f76e4c2c7bd2f027bd049b9a207 Stop aarch64 falling back to SIMDLevel::NONE in faiss dispatch (#5572)

00928e3f7628b8288c17cfece5ee9600a8287f56 Fix ROCm CI: build in a dedicated conda env instead of the runner's broken base (#5545)

1e5780fbbfbe191f3ed3517f93ace8301239e9b2 Fix OOB read in Panorama::reconstruct via IndexPreTransform (T287094917) (#5562)

2ed4c106e9fb9686e7727e5daf8ad6ad1e164109 Fix: Windows SIMD detection (#5497)

b4c66ba6031d1f659163e43430f8ce745b0dbe79 Fix reverse_index_factory/get_code_size round-trip for HNSW, IMI and IVFPQR (#5564)

c0084f5cd2534236acda05cf7b8e0fc8062ddec3 Clamp EDEN unbiased L2 distances to be non-negative (#5569)

6457e941f5f4808303bb480679cb78df0df51934 Contain DirectoryArchiver entries within the extraction root (#5561)

ac60182fc009476c91a2788e639191d1f624157f RaBitQ: Fix zero residual sign convention (#5553)

1a0cfa64e015923d5009d8c7d917aa591e9eac9a Apply the deserialization byte limit to SVS IVF (#5568)

718002f7cf269f3f3752dde04a4cc3f8eabb0050 Fix distance conversion in IndexBinaryFromFloat for non-L2 metrics (#5549)

02ea14372c9983f3eaa15698512180dc62ea234d Fix range_search_max_results to respect all similarity metrics (#5533)

bd7087453e6d2262b6e6c76ab0cfa1e2dfeddebf fix(faiss): use is_similarity_metric for IndexShardsIVF merge (#5541)

561e6ecdf3337e32542b82c735bfaf83d9e3afaa Merge shard results by metric type rather than assuming similarity (#5509)

613e0acac507c7593ff14dee2c6204a83a539acb Fix IndexIDMap range search parameters (#5536)

8b01f308f09cce7f259a1b8978f82c90fcd54c48 Validate IndexScalarQuantizer codes buffer size on deserialization (#5540)

2832ac3f671f83617087789691c79670ee57c728 Fix empty-index BLAS search results (#5528)

e135a1a667a1cfd326faece69d33cf9708a97a7c Fix OSX arm64 nightly: bump libopenblas from 0.3.33 to 0.3.34 (#5538)

920a631851cc7fefe0645c9af81071bb7b2c1237 Don't raise HNSW entry_point to a point that isn't linked yet (#5524)

e2f9ccaccc38b4cc3b65c1c70cb076bb639ef8ca Fix SQtqmse codebook scale: fold 1/sqrt(d) into Lloyd-Max table (#5517)

151452ec94b0df219820b120a365c5396f63bef0 Fix SIMD level detection in Faiss (#5507)

02dbb4e42f49aa2ce57cd78977b0e47a11f7cb91 Fix data race in NNDescent::update() (#5413)

a9ecee221dd3887b3af886b68c7598d8553a5acd Reject AdditiveQuantizer with zero-bit code size on deserialization (#5506)

4de3bbb6017910ecb3db1d3a9086318c517f6e39 Faiss OSS nightly autofix (#5505)

b5ca016ded0921401b532bfd256c2f636406df38 Fix IndexBinaryIVF reconstruct_n and search_and_reconstruct byte stride (#5463)

b3942aba5aa318f39e04d36baa00cb118098dc74 Avoid signed int overflow in VStackInvertedLists binary search (#5482)

Deprecated

a74867cf56a73234ad5d957e40c35c9616d4d273 faiss HNSW: delete the flag and the lock-based graph build (#5611)

5702cafcb75359780cefbc5d50b512ac2e5cfde7 Remove dead TurboQuant projection state (#5618)

1.15.03 Aug 2026

4332ceb Enable byte-LUT gather path for QT_8bit_eden distance computation

[1.15.0] - 2026-07-31

Full changelog

Added

4332ceb Enable byte-LUT gather path for QT_8bit_eden distance computation (#5498)
41133fd Add EDEN quantizer indexes (#5278)
3110fd2 Add MetalIndexIVFPQ with product quantization and residual encoding support (#5288)
6c2b687 Adding cuVS IVF-SQ backend (#5306)
4d778bc Add multi-GPU CAGRA → HNSW build with trainAllNeighbors and multi-GPU optimize (#5282)
42dc286 Implement memory-mapped I/O for Flat and static Vamana SVS indices (#5271)
d22107b feat(riscv): Add RISC-V RVV rabitq kernels (#5369)
df51012 feat(riscv): Add RISC-V RVV vector distance kernels (#5354)
3b52da3 Add AVX2 and AVX512 byte-domain kernels for QT_8bit_direct_signed (#5409)
88b6c47 Implement IVFRaBitQ FastScan conversion constructor (#5422)
d24cc84 Implement GPU IVF add_preassigned via add_core (implement #3908) (#5372)
5826578 Read/Write support for IndexHNSW base class (#5305)
a3350f6 faiss: support METRIC_L1 and extra metrics in IndexFlat range_search (#5391)
432e8b4 Extend DistanceComputer batching to PQ, SQ, and PreTransform (#5309)
7d4bb39 Extract super_kmeans_assign_iteration as a shared public SuperKMeans kernel (#5443)
aff8589 c_api: expose ScalarQuantizer codec and complete FaissQuantizerType enum (#5477)
13aa39b Clone index support for HadamardRotation (#5475)
3e313e4 SIMDConfig::has_dynamic_dispatch() (#5460)
a6f9775 Add faiss-gpu-cuvs pip wheel packaging (#5250)
6227323 pyfaiss_gpu_cuvs_2602: cuVS faiss Python for aarch64/GH200 (26.02, CUDA 12.4)
522644c Stand up parallel cuVS 26.02 IVF-PQ targets (rmm/raft/cuvs) at CUDA 12.8 (#5414)
edebabf Include Python type stubs (init.pyi, py.typed) in conda packages (#5394)
9579d8f Add type stubs for SVS index classes + IndexIVFFlatPanorama.batch_size (#5489)
fd7be0c Add QT_0bit and QT_2bit_tq..QT_5bit_tq to ScalarQuantizer pyi stub (#5323)
ccecc84 Add ClusteringInitMethod enum constants to init.pyi stub (#5324)
55990eb Port test_rabitq_utils from C++ to Python (#5442)

Changed

0236390 Optimize RaBitQ fastscan query setup + 80% qps (#5396)
b1d61c3 Fuse RaBitQ AND-dot and popcount scan (#5412)
81031a9 RaBitQ fastscan avx512 lut quant (#5397)
b7f2b0f RaBitQ SIMD-optimized bit-plane operation (#5392)
0a27700 Reuse RaBitQ IVF scanner distance computer (#5390)
a3d59f0 faiss DD: use 256-bit fast-scan QBS kernel on AMD Zen 4 (split AVX-512) (#5488)
c374dd9 Panorama HNSW Optimizations (~x1.2) (#5190)
cb9a72a Reuse the HNSW visited table across searches instead of reallocating per call (#5448)
4d16574 Raise HNSW visited-table hash-set threshold from 500k to 10M (#5446)
a4b4c0f Optimize exhaustive_L2sqr_blas_cmax with AVX-512 intrinsics. (#4569)
4a2f7e6 Optimize the DistanceComputerByte::compute_code_distance with AVX512F, AVX512BW, AVX512VL instructions (#5067)
9a74c6d Speed up Top1 and Reservoir add_results with AVX2/AVX512 (#5297)
4030faa Speed up polysemous training with AVX-512.
ea10764 Parallelize inverted-list append loop in IndexIVFPQ::add_core_o (#5382)
cf04570 ProductQuantizer::search_sdc: hoist per-subquantizer row pointers out of database-vector loop (#5338)
caf88fc Improve ranklist intersection size (#5237)
5c1afee Remove dead code in quantize_ex_bits to avoid wasted compute (#5384)
e3dd6a8 Remove unused fields in various DistanceComputers (#5318)
35c5495 Remove confusing double-negative throw-macro conditions (#5471)
9223d62 Rename SQI8 to SQ8 to be consistent with the rest of Faiss (#5337)
b37e929 Reuse faiss.ResultHeap in benchs instead of duplicating it (#5363)
1f755ef Prune cuVS brute-force flat index on aarch64 (FAISS_CUVS_NO_FLAT) (#5431)
3a7cce4 Faiss OSS CI: skip GPU jobs on PRs that touch no GPU-relevant code (#5402)
0f48ae6 Faiss OSS autofix: force cuVS nightly to break (E2E validation) (#5328)
0f0f728 Format faiss Python files with black (line length 80) (#5349)
7a7a8f2 Align faiss .flake8 with black and the 80-char standard (#5350)
259fa85 black -l 80: gpu/test formatting (1/3) (#5426)
a7d91d4 black -l 80: tests/ formatting (2/3) (#5427)
3937552 black -l 80: benchs + python + refine test formatting (3/3) (#5428)
b75a23c black -l 80: python/init.pyi formatting (4/4) (#5429)
a238933 Finish E501 line-length lint sweep across faiss (183 violations, 52 files) (#5444)
d4564bd facebook-unused-include-check in distances_avx2.cpp (#5462)
ab3a5a9 facebook-unused-include-check in IndexIVFPQ.cpp (#5461)
e85b9bd facebook-unused-include-check in test_svs.cpp (#5455)
5e1ccce facebook-unused-include-check in TestGpuIndexIVFScalarQuantizer.cpp (#5454)
003e002 facebook-unused-include-check in IndexBinaryHash.cpp (#5453)
c22dc24 facebook-unused-include-check in avx2.cpp (#5452)
76e2bbc facebook-unused-include-check in IndexBinaryHNSW.cpp (#5379)
7c643c2 facebook-unused-include-check in IndexIVFPQFastScan.cpp (#5378)
3c5c27c facebook-unused-include-check in avx2.cpp (#5377)
9b33960 facebook-unused-include-check in test_ivf_early_termination.cpp (#5376)
027054f facebook-unused-include-check in IVFPQ_QueryTables.cpp (#5359)
f4cb963 facebook-unused-include-check in avx2.cpp (#5358)
afc1be4 facebook-unused-include-check in hamming.cpp (#5355)
6256cc7 facebook-unused-include-check in IndexIVF.cpp (#5356)
83dc6a6 facebook-unused-include-check in IndexIVFSpectralHash.cpp (#5357)
124bfa1 facebook-unused-include-check in IndexSVSVamanaLeanVec.cpp (#5343)
a0457db facebook-unused-include-check in avx2.cpp (#5344)
595c5d7 facebook-unused-include-check in distances_simd.cpp (#5342)
5437256 facebook-unused-include-check in avx2.cpp (#5314)
7c56a9d facebook-unused-include-check in avx2.cpp (#5315)
0f280a7 facebook-unused-include-check in distances_simd.cpp (#5316)

Fixed

6187d94 clang-diagnostic-unused-parameter in GpuCloner.cpp (#5480)
37313e7 Fix two memory leaks in shard_ivf_index_centroids (#5495)
7a4e797 Fix RaBitQ fast scan aux offsets for large bbs (#5421)
a032797 faiss DD: SQ AVX-512 falls back to AVX2 when d%16!=0 (#5478)
de8520b missing auto vectorization for AVX512-SPR (#5433)
4a23937 Fix AVX-512 PQ-encode falling back to scalar in fvec_L2sqr_ny_nearest_y_transposed (#5330)
7437cac Fix Jaccard heap selection in IVF search (#5408)
458c7be Fix division-by-zero in quantize_lut when all LUT entries are identical (#5424)
ecd5b65 Guard constant/zero-norm query in RaBitQ query quantization (#5381)
1190b70 RaBitQ: Clamp L2 distance estimate to >=0 (#5393)
af45a3b Throw instead of abort on null buffer in RaBitQuantizer::decode_core (#5465)
5372c88 Guard exhaustive_{inner_product,L2sqr}_seq against empty query batches (nx=0) (#5445)
764b86f Guard ScalarQuantizer compute_codes/decode OMP region for small n (#5423)
19f477a Fix out-of-bounds memory access in bvec_checksum (#5310)
3553e01 Fix IndexAdditiveQuantizer::symmetric_dis() (#5319)
b7cd013 Filter out-of-radius results in IndexRefine::range_search (#5370)
4213672 Always widen IndexRefine::range_search base radius by k_factor (#5395)
ed29dbb faiss: fix IndexIDMap range_search crash with empty SearchParameters (#5388)
828de1b fix IDMap add_sa_codes returning -1 labels with ivf subindex Issue #5333 (#5345)
bf53109 Fix get_code_size to handle HNSW coarse quantizers with any M value (#5416)
e1f8d4f Fix reverse_index_factory for multi-bit RaBitQ indexes (#5419)
7b1d0ca Fix C++ reverse_index_factory: missing qtypes crash telemetry, IDMap2 misidentified (#5417)
cc9be4b fix: reverse_index_factory missing 11 SQ types including all TurboQuant variants (#5322)
350a50b Fix OperatingPoints::add prune loop leaving dominated points at equal time (#5415)
bd67601 Fix OperatingPoints leaving dominated points on the Pareto frontier for equal-time points (#5334)
376930b Fix null dereference, dead condition, and Python binding crash in AutoTune and Clustering (#5321)
e306f30 Fix PyCallbackShardingFunction to properly decrement references (#5389)
d91b73d fix: bs add n fix valid counter skip (#5383)
247527b Fix reconstruct_n stubs resolving to torch.Tensor for numpy calls (#5411)
a3e8c67 Fix heap-buffer-overflow in ResidualCoarseQuantizer deserialization (T280514309) (#5485)
4d74915 Bound lattice_nbit loop in IndexLattice ctor to fix deserialization timeout (T277092535) (#5484)
5987f

Note truncated.

Open source →

Additional notes2 sources agree

Added

4332ceb99492e3ba4ce09aada9bd9e48dc4e2f0b Enable byte-LUT gather path for QT_8bit_eden distance computation (#5498)
41133fde4222a8862650601d85f9af684314ebc8 Add EDEN quantizer indexes (#5278)
3110fd242458bf1458507d9d281847543e3b4d92 Add MetalIndexIVFPQ with product quantization and residual encoding support (#5288)
6c2b68775a21a24870f7fd10788ebc4ec0582650 Adding cuVS IVF-SQ backend (#5306)
4d778bc595917de51a0e76de8351984b8bd8203f Add multi-GPU CAGRA → HNSW build with trainAllNeighbors and multi-GPU optimize (#5282)
42dc2862c293b7c7f788be686cf6c9ac93abf6cf Implement memory-mapped I/O for Flat and static Vamana SVS indices (#5271)
d22107beca7a6db4ff146700b19d507e52b595b1 feat(riscv): Add RISC-V RVV rabitq kernels (#5369)
df51012a606bc352058e95fe58d45b348ffb201a feat(riscv): Add RISC-V RVV vector distance kernels (#5354)
3b52da37c40973d4816ff7d2a4e3fb9f74379988 Add AVX2 and AVX512 byte-domain kernels for QT_8bit_direct_signed (#5409)
88b6c47cf514d3bee7d1995c09a39234fa89d8f9 Implement IVFRaBitQ FastScan conversion constructor (#5422)
d24cc844d2cabce819cda806217b243177293e09 Implement GPU IVF add_preassigned via add_core (implement #3908) (#5372)
5826578a7e8f3f9e45bdb74df0c504c242dfdba0 Read/Write support for IndexHNSW base class (#5305)
a3350f6641f85d1e2d08fa191863bde6be470df9 faiss: support METRIC_L1 and extra metrics in IndexFlat range_search (#5391)
432e8b4248c2778b5023d5e2ac15b264ec120d02 Extend DistanceComputer batching to PQ, SQ, and PreTransform (#5309)
7d4bb39f7eb3e9bb4d160aa38ec821ee1a407afc Extract super_kmeans_assign_iteration as a shared public SuperKMeans kernel (#5443)
aff8589c9e1c2d6ee6a0de615aa719a9a4fd6e86 c_api: expose ScalarQuantizer codec and complete FaissQuantizerType enum (#5477)
13aa39b7916824bc09ca6af7b5e241c2cba68f6d Clone index support for HadamardRotation (#5475)
3e313e4c496b6a042b71193fe0bd9dd713efc7a2 SIMDConfig::has_dynamic_dispatch() (#5460)
a6f97750aeaf760c72d0526c3a346513823c4a0c Add faiss-gpu-cuvs pip wheel packaging (#5250)
6227323ffd0341679299ba9a8e71cac1e15674ee pyfaiss_gpu_cuvs_2602: cuVS faiss Python for aarch64/GH200 (26.02, CUDA 12.4)
522644cb2b7a29737fe669b263eccb946217dc93 Stand up parallel cuVS 26.02 IVF-PQ targets (rmm/raft/cuvs) at CUDA 12.8 (#5414)
edebabf90fe79829ede6eaa02f3d9a78d3008a5b Include Python type stubs (init.pyi, py.typed) in conda packages (#5394)
9579d8f0a4dbc91bd7ce195ef0aceb937c89159e Add type stubs for SVS index classes + IndexIVFFlatPanorama.batch_size (#5489)
fd7be0cfb0c8cee0d475bb6285ad996a87ebf917 Add QT_0bit and QT_2bit_tq..QT_5bit_tq to ScalarQuantizer pyi stub (#5323)
ccecc847db5e1d20e929cf8df4a8ddb654769890 Add ClusteringInitMethod enum constants to init.pyi stub (#5324)
55990eb91f9042d811a76d72e43f9ca126d77beb Port test_rabitq_utils from C++ to Python (#5442)

Changed

0236390f2fd8e63ff5e4c6d4ca981b7ba032cd17 Optimize RaBitQ fastscan query setup + 80% qps (#5396)
b1d61c3b883fb82063e024979404dd6ebe134eca Fuse RaBitQ AND-dot and popcount scan (#5412)
81031a97892896ca9167e5203e27420de4617e67 RaBitQ fastscan avx512 lut quant (#5397)
b7f2b0f1fd8d62e902bdcdcb9ae9694a6fd8f792 RaBitQ SIMD-optimized bit-plane operation (#5392)
0a27700e05429e60221b0eccbf50c4a79ad2e450 Reuse RaBitQ IVF scanner distance computer (#5390)
a3d59f0999c69015eeff54fdde3e0d051798179e faiss DD: use 256-bit fast-scan QBS kernel on AMD Zen 4 (split AVX-512) (#5488)
c374dd917b75c8b1fef56e487d26e18d3aae5605 Panorama HNSW Optimizations (~x1.2) (#5190)
cb9a72ac03f4e7f7d563e3e54ab4e86b8ce1af12 Reuse the HNSW visited table across searches instead of reallocating per call (#5448)
4d1657471f8590ccadf5c2f37d6b58999b8010ec Raise HNSW visited-table hash-set threshold from 500k to 10M (#5446)
a4b4c0f2de14d30719e0750e40922e47517eaa80 Optimize exhaustive_L2sqr_blas_cmax with AVX-512 intrinsics. (#4569)
4a2f7e63280ec095bb16dbf0d8e837b2ab79c27f Optimize the DistanceComputerByte::compute_code_distance with AVX512F, AVX512BW, AVX512VL instructions (#5067)
9a74c6d3d14ccf02157782c8f47205765fad63d9 Speed up Top1 and Reservoir add_results with AVX2/AVX512 (#5297)
4030faae3d5b2c67a5931a637a1e77e2ceee9e1e Speed up polysemous training with AVX-512.
ea107641d4ccc350e9c3eebaffcada74b586d801 Parallelize inverted-list append loop in IndexIVFPQ::add_core_o (#5382)
cf045707c5342cf670b9e1176d47337b05fa0b6e ProductQuantizer::search_sdc: hoist per-subquantizer row pointers out of database-vector loop (#5338)
caf88fca98d6693366642cbb7d9b8df27dcb91ce Improve ranklist intersection size (#5237)
5c1afeed9666448708c8a036136a8a8bdf8827e1 Remove dead code in quantize_ex_bits to avoid wasted compute (#5384)
e3dd6a85f3125d3338b2a5918e007230927c929f Remove unused fields in various DistanceComputers (#5318)
35c5495b8fa8fcab4b8b476d6a6efde67d3a0e85 Remove confusing double-negative throw-macro conditions (#5471)
9223d62e81d8cc5e0b5fb9aa20c48356c9670968 Rename SQI8 to SQ8 to be consistent with the rest of Faiss (#5337)
b37e929947d9116140392bc84bc6f0fea0c36b00 Reuse faiss.ResultHeap in benchs instead of duplicating it (#5363)
1f755ef3d6bd274f64d344618d5c98dccd36d962 Prune cuVS brute-force flat index on aarch64 (FAISS_CUVS_NO_FLAT) (#5431)
3a7cce4ea9861c14fb50c1b18ffbdd012312cda3 Faiss OSS CI: skip GPU jobs on PRs that touch no GPU-relevant code (#5402)
0f48ae62101052cd6874381bc455155390891cad Faiss OSS autofix: force cuVS nightly to break (E2E validation) (#5328)
0f0f728166218eab2605054c713cdf2646628014 Format faiss Python files with black (line length 80) (#5349)
7a7a8f253207f684bbe0bac763e723f341f83b3e Align faiss .flake8 with black and the 80-char standard (#5350)
259fa853af5974186797fc135ac0bd2b71f0b92b black -l 80: gpu/test formatting (1/3) (#5426)
a7d91d44d945b58baebe03ce4fdbdfecc55dec73 black -l 80: tests/ formatting (2/3) (#5427)
3937552dd2e7309158804b024c68c7d0910761c4 black -l 80: benchs + python + refine test formatting (3/3) (#5428)
b75a23ca15996d9e33115b0c8730be1e30a0f515 black -l 80: python/init.pyi formatting (4/4) (#5429)
a238933d17b9abd0c0137cc7d0581bde7f164b46 Finish E501 line-length lint sweep across faiss (183 violations, 52 files) (#5444)
d4564bdeb2fcecb30a4957a7479be35c71b2d364 facebook-unused-include-check in distances_avx2.cpp (#5462)
ab3a5a95a61f42713cca0c15e67a290c8bcfcef5 facebook-unused-include-check in IndexIVFPQ.cpp (#5461)
e85b9bded90f06f9543e71e8c06686683dba8224 facebook-unused-include-check in test_svs.cpp (#5455)
5e1cccea390984f52994d256c82c27c2e398fffe facebook-unused-include-check in TestGpuIndexIVFScalarQuantizer.cpp (#5454)
003e002da66f523fd01254164677481f94025e22 facebook-unused-include-check in IndexBinaryHash.cpp (#5453)
c22dc247c095470580cf35b2be6b608976e24ea2 facebook-unused-include-check in avx2.cpp (#5452)
76e2bbca43a5babdc8d7ca9b4e3ca175ffc39bfa facebook-unused-include-check in IndexBinaryHNSW.cpp (#5379)
7c643c293067bc682085a8b16eba7b936316d265 facebook-unused-include-check in IndexIVFPQFastScan.cpp (#5378)
3c5c27c881131d28794d54472ee18d9ce6a8e75f facebook-unused-include-check in avx2.cpp (#5377)
9b33960463a3ffdbfa1aaf45325d5258b607cbd7 facebook-unused-include-check in test_ivf_early_termination.cpp (#5376)
027054f40f78c4a8a42eeea43561ab99fb5deaf5 facebook-unused-include-check in IVFPQ_QueryTables.cpp (#5359)
f4cb963381c8afc3e5ba749df692fc017a7c17ff facebook-unused-include-check in avx2.cpp (#5358)
afc1be442b5f48082a346f78b6d9e0402862cac4 facebook-unused-include-check in hamming.cpp (#5355)
6256cc7bd748ca8dd0c09d2b43b7ff52a84fb124 facebook-unused-include-check in IndexIVF.cpp (#5356)
83dc6a6991abc67bad2026dbca94bb6696cc3cdf facebook-unused-include-check in IndexIVFSpectralHash.cpp (#5357)
124bfa1d4f9f8f3b114bf9da602941d815248e07 facebook-unused-include-check in IndexSVSVamanaLeanVec.cpp (#5343)
a0457db86cd89e515782f65980af2748990184fb facebook-unused-include-check in avx2.cpp (#5344)
595c5d7aa1f8a2f23a228faa8a9ab05cb6fe3acd facebook-unused-include-check in distances_simd.cpp (#5342)
5437256fc557d86fd9a251ff1436b65eda10c203 facebook-unused-include-check in avx2.cpp (#5314)
7c56a9dacd8ca2a4b161d57050e56a8222eac650 facebook-unused-include-check in avx2.cpp (#5315)
0f280a7ca93da5ce9da6944ace8c1ff62afd3c18 facebook-unused-include-check in distances_simd.cpp (#5316)

Fixed

6187d940e329ae9d6073fe9d4c429a385707f606 clang-diagnostic-unused-parameter in GpuCloner.cpp (#5480)
37313e7b6487544e899afce9338f3f987361222f Fix two memory leaks in shard_ivf_index_centroids (#5495)
7a4e7975d1aa88a783e1304932346e3dce7214d1 Fix RaBitQ fast scan aux offsets for large bbs (#5421)
a032797bd8417a4003f6ceefc2456a7368c1bb7c faiss DD: SQ AVX-512 falls back to AVX2 when d%16!=0 (#5478)
de8520b7e45b3f8e5a2d8d35bd00d6602773501b missing auto vectorization for AVX512-SPR (#5433)
4a23937417af3ae41dfdf891773e8e81283ae986 Fix AVX-512 PQ-encode falling back to scalar in fvec_L2sqr_ny_nearest_y_transposed (#5330)
7437cac4d9ad071cc76dcde570b3080cbb13f9d1 Fix Jaccard heap selection in IVF search (#5408)
458c7be73520357ce6485fcc3e33d58ba09b4457 Fix division-by-zero in quantize_lut when all LUT entries are identical (#5424)
ecd5b65dbab001e8953c41013650ef1ad90fbf21 Guard constant/zero-norm query in RaBitQ query quantization (#5381)
1190b707c2c7b59e371d9a2eda73338f816c07c0 RaBitQ: Clamp L2 distance estimate to >=0 (#5393)
af45a3b2a706188d4f3fa8543f49e6e4285558f6 Throw instead of abort on null buffer in RaBitQuantizer::decode_core (#5465)
5372c889897ef29d003823aa73d9e5e7459f9680 Guard exhaustive_{inner_product,L2sqr}_seq against empty query batches (nx=0) (#5445)
764b86f24cddd529af5e3cdb779f5c4a95483a31 Guard ScalarQuantizer compute_codes/decode OMP region for small n (#5423)
19f477a9ba2cb25b62aed8d6dd10c4b7e66ca337 Fix out-of-bounds memory access in bvec_checksum (#5310)
3553e01cd9269917dfb27204c5781ccc24c73130 Fix IndexAdditiveQuantizer::symmetric_dis() (#5319)
b7cd013065b41d8f3b1a303591f0e64bec14eb70 Filter out-of-radius results in IndexRefine::range_search (#5370)
4213672cb8c7f4c5fff2e0db50afb739ce1927c1 Always widen IndexRefine::range_search base radius by k_factor (#5395)
ed29dbb0bbe38b19bea0649b0b4f85d2d43eec06 faiss: fix IndexIDMap range_search crash with empty SearchParameters (#5388)
828de1b77eb356899b91b6a1fb603a4c31802336 fix IDMap add_sa_codes returning -1 labels with ivf subindex Issue #5333 (#5345)
bf53109ec1a1124be62d7d346400afd05c0d523c Fix get_code_size to handle HNSW coarse quantizers with any M value (#5416)
e1f8d4fc0d99873b42485c5333dd6cb16d958794 Fix reverse_index_factory for multi-bit RaBitQ indexes (#5419)
7b1d0ca4355b615e9a46d8eb79b443fb20784b1b Fix C++ reverse_index_factory: missing qtypes crash telemetry, IDMap2 misidentified (#5417)
cc9be4b1e42c814655c1a4a3143a95db9684ac3c fix: reverse_index_factory missing 11 SQ types including all TurboQuant variants (#5322)
350a50b02ede0e5bdeca3fe178fde1616828df37 Fix OperatingPoints::add prune loop leaving dominated points at equal time (#5415)
bd6760153c4554bf257ec66af528f1546284aa9b Fix OperatingPoints leaving dominated points on the Pareto frontier for equal-time points (#5334)
376930beabd2c0e24eb86f4e1209f6594f750b9d Fix null dereference, dead condition, and Python binding crash in AutoTune and Clustering (#5321)
e306f3036d326f70797588b5f1eaea5f340959df Fix PyCallbackShardingFunction to properly decrement references (#5389)
d91b73dc63059fa98a29c6f44d330bedaed4dd49 fix: bs add n fix valid counter skip (#5383)
247527b3490d5a4019ecb4aa6062de71dce1a81a Fix reconstruct_n stubs resolving to torch.Tensor for numpy calls (#5411)
a3e8c67c60a7ba9dcca511f782deef26c11f252f Fix heap-buffer-overflow in ResidualCoarseQuantizer deserialization (T280514309) (#5485)
4d7491540d5404c062d4f3210e6fd0c177388af8 Bound lattice_nbit loop in IndexLattice ctor to fix deserialization timeout (T277092535) (#5484)
5987f523f491eca461a3fec1bb26258480d3bd76 Reject nbits==0 ProductQuantizer on deserialization (T279464247) (#5483)
269ac43d618bd9169f36806b89155d31ac31994b Bound recursion depth in index deserialization (CWE-674) (#5468)
a20b22d0d45bd0cc723449710a15b5d23d869fc1 Bound decode-cache compute in ZnSphereCodecRec deserialization (#5467)
6a0616edbda5a177011af59d2cc1021fb7054aa0 Reject null base/refine index in IndexRefine deserialization (#5466)
bc102d61b8a71164f3dc53e2eaabb19ef7ea524c Bound IVFPQ precomputed-table allocation at deserialization time (#5464)
756c5d5773f65a9477cdc2e369b3819b503d0c7e Harden FastScan deserialization: enforce ksub==16 and bound reconstruct() (#5436)
f1efc2cf89bd4d0f2c5317808b3c43d0694856dd Validate codes/ntotal consistency in Ix2L and IxLa deserialization (#5439)
fba9084a62dd443d289bac79bbe3c5a47584b2f9 Validate Panorama batch_size and codes/cum_sums on deserialization (#5438)
d43336302e06aec21bd0822ef0cbb01c8c34d6be Validate SVS-Vamana stored_vectors size on deserialization (ISV2) (#5434)
dfd2a25aed4de7959a4740aa18e70f29be82f38b Reject degenerate code_size==0 IndexPQ with ntotal>0 on deserialization (#5437)
91ce5f8a9f3df04e57b34b96071de65e1605c371 Harden reconstruct()/reset() against null storage in composite indexes (#5440)
ea8e22c1d3d31195683a18ebf224d0b94c8597cb Bounds-check decoded direct_map entry in IndexIVF::reconstruct (#5435)
79f59edcfc17f532b69fd37a1887751e4ad6c3ef Validate legacy IVFFlat per-list codes size against ids size (#5259)
c428ca1e3886620343fe86c322c709b9b787ddc5 Harden OnDiskInvertedLists mmap reader against malformed list sizes (#5312)
6f62d01afa4feb36b69f66e7bcabc959c49880b1 Guard against integer overflow in index deserialization storage-size checks (#5313)
16c5d782e0d862b915860c1b0b734189105a3416 Don't redefine PRETTY_FUNCTION under clang-cl (#5474)
a9683455b757bbffd913630806fc80ce07f65c75 clang-diagnostic-unused-parameter in bench_result_handler_overhead.cpp (#5481)
d6d3dce8d7fb16b9b1fc1c75e45a6e8337e08f08 clang-diagnostic-unused-parameter in bench_cppcontrib_sa_decode.cpp (#5479)
8d0d1a2b3345646b959bfb4962f4a3061aecc3d7 faiss/gpu/perf: check CUDA return in PerfIVFPQAdd (clang21 nodiscard) (#5368)
0fa5b0afcbb8f2318a70fccf209491723df64ff6 faiss/gpu/perf: check cudaProfilerStop() return (fix clang21 -Wunused-value) (#5362)
6f0155705b0281e6601c3a6f4fb610de0889289a Keep distances_simdlib256.h include in distances_simd.cpp (unused-include false positive) (#5380)
22220d2c4f45ee3b7fa864ed0e764bdf336dff7c Fix genuine flake8 lint in faiss Python (F401/F403/F841/B011) (#5351)
63378f0ace9771397e6d3474e66317fb96a97c71 Fix cuVS GPU CI timeout: pin CUDA toolkit to 13.2 (#5401)
11bd76463c6e0227668a7d4cae03a2ed5001c0be Faiss OSS nightly autofix (#5346)
7cd1ae906c126d8a02721831a47068f1f3e695de Fix panorama tests: pass D_panorama (not D_regular) to check_ref_knn_with_draws (#5430)
ba6af9ffe879a4b9353a4a72307a4aca71d21ceb Fix flaky IVF/Refine Panorama kNN tests under AVX512_SPR dispatch (#5420)
6653b98402b9fae611033d1fddc35736622526c0 Fix flaky TestIVFRaBitQ_AVX2::test_comparison_vs_ref_L2_rrot — apply cross-SIMD equivalence (#5325)
aa9d2bf4256da7ca8cc36d743fda53ad392aae59 De-flake test_flat_panorama across SIMD levels (#5361)
395eaf8c6f29efb4aeb1f3bb3540aeecbb464d98 De-flake panorama / graph_based / hamming_utils tests (#5339)
08cb1f2cf32675c3c8dc9dd7460d6fe0d9183dd1 De-flake test_refine_panorama.test_incremental_add (seed + tolerance) (#5347)
8bfcb4e2a31615e2e3740e454c2cd782dc60fc10 De-flake test_meta_index.Shards.test_shards (loosen tie-churn tolerance) (#5348)
54f83a413183659d45e444bf44307903d042037a De-flake bench_transposed_encode (CI timeout) (#5360)

1.14.313 Jun 2026

Metal GPU backend expansion — new MetalIndexIVFFlat with IVF scan/merge kernels and expanded top-k support ( #5202 ); Metal now enabled by default on

Highlights

Metal GPU backend expansion — new MetalIndexIVFFlat with IVF scan/merge kernels and expanded top-k support (#5202); Metal now enabled by default on Apple Silicon machines (#5280)
TurboQuant in ScalarQuantizer — full Algorithm 2 (QJL stage) with SIMD and optimizations (#5170)
Sapphire Rapids (SPR) optimizations — ScalarQuantizer L2/IP (#5173), VPOPCNTDQ-based HammingComputer (#5183), and VPOPCNTDQ-based RaBitQ kernel (#5149)
faiss-gpu pip wheels — new GPU wheel packaging (#5131); musllinux wheels re-enabled for faiss-cpu (#5299)
SVS static vamana support (#5224)
HNSW is_similarity mode for IP/similarity metrics (#5246, #5226)
HNSW search performance — reserve VisitedTableSet capacity to avoid rehashes (#5290), search_from_candidate_unbounded templatized for VisitedTable devirtualization (#5270), and runtime checks avoided in VisitedTable (#5234)
NEON FINE_SIZE=2 specializations for Index2LevelDecoderImpl and IndexPQDecoder (#5255)
cuVS upgraded to 26.06 (#5240); CI updated to ROCm 7 (#5196)
Deserialization hardening — bool-field validation (#5279) and null inner-index rejection in IDMap / BinaryFromFloat (#5239)
Build robustness — fix non-AVX2 import SIGILL via .rodata partitioning tables (#5298), Windows ARM64 (MSVC) NEON build fix (#5274), and AVX512_SPR dispatch fix on AMD (#5281)

Full Changelog

Added

261d8f2 Add IVFSQTurboQSearchParameters to init.pyi stub (#5304)
684a32d Add manual Faiss nightly workflow dispatch (#5300)
10b6b2a Add MetalIndexIVFFlat with IVF scan/merge kernels and expanded top-k support (#5202)
74d3619 TurboQuant in ScalarQuantizer: full Algorithm 2 (QJL stage) with SIMD and optimizations (#5170)
5066009 Add Python HNSW tutorial (#5260)
3cff0f4 Add Sapphire Rapids optimizations for ScalarQuantizer (L2, IP) (#5173)
46ef80d Support IndexIDMap/IndexIDMap2 in reverse_index_factory (#5266)
420158b Add VPOPCNTDQ-based HammingComputer for Sapphire Rapids+ (#5183)
d8f1c27 Implement NEON-based FINE_SIZE=2 specializations for Index2LevelDecoderImpl and IndexPQDecoder (#5262)
0951b53 Support user provided blas library (#5189)
215740e SVS static vamana support (#5224)
d24ad6e Add is_similarity mode to HNSW (#5246)
aa332cd Implement NEON-based FINE_SIZE=2 specializations for Index2LevelDecoderImpl and IndexPQDecoder (#5255)
5d9aae6 Add faiss-gpu pip wheel packaging (#5131)
6f1cf64 Add VPOPCNTDQ-based RaBitQ kernel for Sapphire Rapids+ (#5149)

Changed

262fc3c Re-enable musllinux wheels for faiss-cpu (#5299)
1cdc370 Run CI on push to main to refresh ccache cache (#5291)
379ee75 Reserve VisitedTableSet capacity to avoid rehashes during HNSW search (#5290)
6513a24 Enable Metal by default on Apple machines (#5280)
fe46c3c Validate bool fields during deserialization (#5279)
480f917 Type imbalance_factor and wire the .pyi stub into the buck build (#5269)
e60baeb Templatize search_from_candidate_unbounded for VisitedTable devirtualization (#5270)
c000190 Accelerate ScalarQuantizer::QT_bf16 with AVX512-BF16. (#4889)
7504fc8 Upgrade CUVS Version to 26.06 (#5240)
d12683c facebook-unused-include-check in IndexBinaryIVF.cpp (#5263)
99d9013 facebook-unused-include-check in distances_simd.cpp (#5264)
2f0368b facebook-unused-include-check in hamming_avx2.cpp (#5265)
0c72755 Remove unused include of platform_macros.h in partitioning.cpp
e6b8f6d IndexHNSW: use HNSW::is_similarity for IP/similarity metrics + tests (#5226)
c0575f2 avoid runtime checks in VisitedTable (#5234)
1cb7601 Eliminate per-code denormalization in uniform SQ distance computation (#5166)
f29d862 Revert D106693266: Implement NEON-based FINE_SIZE=2 specializations for Index2LevelDecoderImpl and IndexPQDecoder
ef96e3d Updating CI to ROCm 7 (#5196)
f5217d7 facebook-unused-include-check in IndexBinaryHNSW.cpp (#5251)
3e6ed99 facebook-unused-include-check in IndexBinaryIVF.cpp (#5252)
108868b Reject null inner index in IDMap and BinaryFromFloat deserialization (#5239)
a64b549 Add IndexLattice r2 limit to cap decode-cache build cost (#5238)
0993715 faiss: Replace remaining get_single_code calls with ScopedCodes (#5248)
910e435 facebook-unused-include-check in hamming_avx2.cpp (#5242)
d581f2f Use per-SIMD TU scan for standalone PQ (AVX2 gather inlining) (#5233)

Fixed

20afed0 make intentional cudaGetLastError() error-clears explicit ((void)) for clang21 -- fixes S674096 (#5302)
e420e94 Move partitioning shifts tables to .rodata to fix non-AVX2 import SIGILL (#5298)
15bd823 Fix cuVS nightly (#5273)
17cc967 Fix AVX512_SPR dispatch on AMD: require AVX512_FP16 CPUID (#5281)
e69bfee Stabilize RaBitQ tests on AVX512_SPR by switching to cross-level equivalence (#5277)
c4c6514 Stub fixes: knn torch overload, ResidualCoarseQuantizer ctor, remove duplicate I/O defs (#5283)
eb4c1ea Fix Windows ARM64 (MSVC) build broken by NEON SIMD templatization (#5274)
ab63238 Install missing InvertedListScannerStats.h header (#5268)
1405415 Fix broken fbcode//faiss/tests:test_index_binary - test_replicas (test_index_binary.TestReplicasAndS (#5258)
34eb989 fix(python): int64 coercions for MapLong2Long + InvertedLists DOWNCAST chain (#5257)
d492af4 faiss: initialize ids_tab to -1 in Top1BlockResultHandler::begin_multiple (#5249)
a4e417f Fix: static SIMD dispatch falls to scalar for avx512_spr/avx512/arm_sve builds (#5057)
3edc6e1 Guard Panorama autovec pragmas against nvcc frontend (#5241)
c7689c4 Fix ODR violation in ScannerMixIn by adding SIMDLevel template parameter (#5148)
ca87f41 Open SIFT demo data in binary mode (#5213)

Full diff: v1.14.2...v1.14.3

Open source →

Additional notes2 sources agree

Added

261d8f25aac03c980f1faf5df775248eff96cb6d Add IVFSQTurboQSearchParameters to init.pyi stub (#5304)
684a32d0404758bddd4f7fb9ec76db54252819e0 Add manual Faiss nightly workflow dispatch (#5300)
10b6b2a1e554afc4cbd66fe73d864728ff9e310c Add MetalIndexIVFFlat with IVF scan/merge kernels and expanded top-k support (#5202)
74d36198faac528fce715aa8731bf589727bef6d TurboQuant in ScalarQuantizer: full Algorithm 2 (QJL stage) with SIMD and optimizations (#5170)
506600962f8a58926710edd24715bc9002eb18e9 Add Python HNSW tutorial (#5260)
3cff0f419c459bdc7c4f56ea49effb6d7da1475c Add Sapphire Rapids optimizations for ScalarQuantizer (L2, IP) (#5173)
46ef80db639e4de03720e3c404065636846ed47e Support IndexIDMap/IndexIDMap2 in reverse_index_factory (#5266)
420158b10e0136228ff634d464fa90c4370a1ead Add VPOPCNTDQ-based HammingComputer for Sapphire Rapids+ (#5183)
d8f1c2716d9c2c070c0d7e4e0c1017331e9e9e8a Implement NEON-based FINE_SIZE=2 specializations for Index2LevelDecoderImpl and IndexPQDecoder (#5262)
0951b5337b72755ade5c02705396fb7d7cedceac Support user provided blas library (#5189)
215740ecc3add92d7c27e24436b2dffd8fdb157c SVS static vamana support (#5224)
d24ad6ef5b3e40c0e4cf8b695923cbcd674d2573 Add is_similarity mode to HNSW (#5246)
aa332cd5f0d6984075c849b32c387fffe067a49a Implement NEON-based FINE_SIZE=2 specializations for Index2LevelDecoderImpl and IndexPQDecoder (#5255)
5d9aae6ee81078d33ff3c4e2fc97f7583e2e450d Add faiss-gpu pip wheel packaging (#5131)
6f1cf64531130db04843737d1f538999fc295583 Add VPOPCNTDQ-based RaBitQ kernel for Sapphire Rapids+ (#5149)

Changed

262fc3c7a2ec099cdd0cd21482931702171c60eb Re-enable musllinux wheels for faiss-cpu (#5299)
1cdc3709c656eea41ce3725024070f2516e3b30b Run CI on push to main to refresh ccache cache (#5291)
379ee758b1bb49cd167c6ed86e80be331e94c593 Reserve VisitedTableSet capacity to avoid rehashes during HNSW search (#5290)
6513a2497cbc29f5c0c74c0d17bbe8b9f6ff3230 Enable Metal by default on Apple machines (#5280)
fe46c3c804296cc04238c2e17b02c671887791a8 Validate bool fields during deserialization (#5279)
480f91790080c8daa92fcb1941c5e8d0f56de237 Type imbalance_factor and wire the .pyi stub into the buck build (#5269)
e60baeb7b61e617a483310fccb239ba204458c7b Templatize search_from_candidate_unbounded for VisitedTable devirtualization (#5270)
c000190dc07c161898093a72a16f3918bd9fd621 Accelerate ScalarQuantizer::QT_bf16 with AVX512-BF16. (#4889)
7504fc8ddebdade7fcabd9dacd9861318d475b50 Upgrade CUVS Version to 26.06 (#5240)
d12683c3b00247ca6333ac717f3a8af49aa835e3 facebook-unused-include-check in IndexBinaryIVF.cpp (#5263)
99d9013e1fbe232356759d3abedbe20df0e8dba1 facebook-unused-include-check in distances_simd.cpp (#5264)
2f0368b6205c92d910ae772417e71b68dde216aa facebook-unused-include-check in hamming_avx2.cpp (#5265)
0c72755eccff39f903d58ed1f46d6c278a914cae Remove unused include of platform_macros.h in partitioning.cpp
e6b8f6de5380aed85c2450abd31ecdb749770bbe IndexHNSW: use HNSW::is_similarity for IP/similarity metrics + tests (#5226)
c0575f2c1319a702fa4ca615e2354014c42b3e67 avoid runtime checks in VisitedTable (#5234)
1cb760182917697677fe100f8a1cc2b4d0b9f751 Eliminate per-code denormalization in uniform SQ distance computation (#5166)
f29d8621fa4250b381d1bfb1d8362c4867ee3de3 Revert D106693266: Implement NEON-based FINE_SIZE=2 specializations for Index2LevelDecoderImpl and IndexPQDecoder
ef96e3d25ecbfaa52894b0f361e9d156ee92254a Updating CI to ROCm 7 (#5196)
f5217d74960ead3a1d2c0b69baa63479b6f49ea2 facebook-unused-include-check in IndexBinaryHNSW.cpp (#5251)
3e6ed99bbb6cfb8daf98cadf904da449a4aaea48 facebook-unused-include-check in IndexBinaryIVF.cpp (#5252)
108868bfac5ad7e0ce1b4e46857d4724a62a309e Reject null inner index in IDMap and BinaryFromFloat deserialization (#5239)
a64b5496762a93cb7836d013c2283c6da479f717 Add IndexLattice r2 limit to cap decode-cache build cost (#5238)
09937153dc55f64321ed7d96730fc3d673a9092d faiss: Replace remaining get_single_code calls with ScopedCodes (#5248)
910e435bccafedfac5aa69f87863085ce6d90a3c facebook-unused-include-check in hamming_avx2.cpp (#5242)
d581f2f8406f8794f58794124b650ab7402ac848 Use per-SIMD TU scan for standalone PQ (AVX2 gather inlining) (#5233)

Fixed

20afed0fa3aefe2d4f5bf6a5f6ab1d1459efec05 make intentional cudaGetLastError() error-clears explicit ((void)) for clang21 -- fixes S674096 (#5302)
e420e94ce694213a8338ac975d565f81d7251d96 Move partitioning shifts tables to .rodata to fix non-AVX2 import SIGILL (#5298)
15bd8238375a1650156dc6e4b8ea54def3a3dc23 Fix cuVS nightly (#5273)
17cc967d711c2647495370a09469e3a710b527ae Fix AVX512_SPR dispatch on AMD: require AVX512_FP16 CPUID (#5281)
e69bfeed087fd70a4708d4c92c4e0bf5e5b888a6 Stabilize RaBitQ tests on AVX512_SPR by switching to cross-level equivalence (#5277)
c4c651453b5a7d73919458c6209573c4c11ddadc Stub fixes: knn torch overload, ResidualCoarseQuantizer ctor, remove duplicate I/O defs (#5283)
eb4c1eafe2f2e9d77d0fc152e331f5a0c8d77e13 Fix Windows ARM64 (MSVC) build broken by NEON SIMD templatization (#5274)
ab632382d3d0531fc6c744e0948544aa358bc850 Install missing InvertedListScannerStats.h header (#5268)
140541590e9beac7f3d81d3dce79af9c42777906 Fix broken fbcode//faiss/tests:test_index_binary - test_replicas (test_index_binary.TestReplicasAndS (#5258)
34eb98936f5c0079ec4e65c712c0c2f38931c80e fix(python): int64 coercions for MapLong2Long + InvertedLists DOWNCAST chain (#5257)
d492af446e847a59bcfa6e35e42f1be1318c40fa faiss: initialize ids_tab to -1 in Top1BlockResultHandler::begin_multiple (#5249)
a4e417fed77705e7918458c8b1caad5efbd506ce Fix: static SIMD dispatch falls to scalar for avx512_spr/avx512/arm_sve builds (#5057)
3edc6e164973f619589ef4c7ea0cce5443e34ecf Guard Panorama autovec pragmas against nvcc frontend (#5241)
c7689c4d0c4e71eed1e43639e5f9973f89a016b9 Fix ODR violation in ScannerMixIn by adding SIMDLevel template parameter (#5148)
ca87f41dfe58336245a013118e38ddaad0f7cab3 Open SIFT demo data in binary mode (#5213)

1.14.222 May 2026Deprecated

582246b Remove use of __AMDGCN_WAVEFRONT_SIZE macro with it being deprecated in ROCm 7

Highlights

SuperKMeans — faster k-means via ADSampling + PDX progressive pruning (#5168)
Metal GPU backend for Apple Silicon, starting with IndexFlat (#5144)
RVV (RISC-V Vector) SIMD backend + RISC-V cross-compilation CI (#5156, #5184)
pip install support via scikit-build-core + cibuildwheel (#4862)
SVS IVF index enabled; SVS upgraded to 0.3.0 (#4801, #4999); SVSVamana usable as IVF coarse quantizer via index_factory (#5175)
TurboQuant (CPU) quantizer (#5049)
HNSW improvements — optional persistent locks for incremental adds (#5031), prune_headroom to avoid O(n²) pruning (#4847), narrower add-time critical section (#4915)
Extensive index deserialization hardening — validation and OOM/overflow protection across ~40 PRs (binary, IVF, HNSW, FastScan, ScalarQuantizer, AdditiveQuantizer, ProductQuantizer, SVS, VectorTransform, …)
Major Dynamic Dispatch refactor across distances, hamming, partitioning, RaBitQ, FastScan; dynamic dispatch now builds on Windows (#5127)
cuVS upgraded to 26.02 (#4945, #4788); CI moved to CUDA 13.2 (#5207); C++20 for cuVS (#4881); filtered search for cuVS indexes (#4858)
OMP exception safety across IndexHNSW, IndexNNDescent, IndexNSG, IndexFlatCodes, IndexIVF (#5106, #5105, #5111, #5133, #5037, #5053)

Full Changelog

Added

b7618fa Add SuperKMeans: faster k-means via ADSampling+PDX progressive pruning (#5168)
66c9d08 Add SuperKMeans foundations: math primitives, PDX layout, SIMD kernels
dc7afcf Support SVSVamana as IVF coarse quantizer via index_factory (#5175)
7a8e4dd CI: cross-compile for riscv64 with RVV dynamic dispatch (#5184)
0320279 Introduce RVV (#5156)
417c53e Add NoneSIMDLevel context manager for cross-level reference checks (#5158)
46def46 Enable SVS IVF Index in FAISS (#4801)
585ba79 Add pip install support via scikit-build-core + cibuildwheel (#4862)
799bf3a Introduce early stop facilities for IVF, attempt 2 (#5160)
66cea52 Add Metal GPU backend for Apple Silicon (IndexFlat) (#5144)
17fd333 Add database-parallel flat search for few-query workloads (#5000)
c7f07bb Add C++ per-SIMD-level tests for distance utility functions (#5154)
0242758 Add bit-exact RQ encode cross-level test (#5155)
2f68a37 Add per-SIMD-level tests for hamming.h utils + fix crosshamming bug (#5153)
a7116ed Add TurboQuant (CPU) (#5049)
11e5d39 Add optional persistent locks to IndexHNSW for incremental adds (#5031)
e82aa28 Add OMP exception capture helpers and migrate IndexIVF (#5111)
6299bff Add QT_0bit to ScalarQuantizer for centroid-only IVF distance (#5079)
3030fe0 Add filtered search for cuVS indexes (#4858)
f83ea51 Support IVFRaBitQSearchParameters in RaBitQFastScan scanner (#5081)
82a8235 Add iterator callbacks for distance computation and heap admission (#5082)
b068fd9 Support IO_FLAG_MMAP_IFC on Darwin (#5058)
113bf6c Expose fast_scan_code_size from non-IVF FastScan indices (#5077)
786d993 Expose a fast scan code size from fast scan indices (#5064)
95f1e45 OMP build diagnostics (#4842)
c4ac852 Add a balanced assignment function (#5050)
8402a6b Add unit tests for HNSW deserialization validation (#5033)
1dc5e1f V2 enable specifying a subset of SIMD levels that are implemented at dispatching time. (#4959)
e6f5c0c Add SIMD dynamic dispatch migration guide (#4973)
8177a3b Add nvidia-smi diagnostic call (#4960)
67fc664 Add CappedInvertedLists for per-list size limiting (#4880)
e57a893 introduce SingleQueryResultCollectHandler (#4926)
748c031 HNSW: add prune_headroom to avoid O(n^2) pruning/locking, headroom test (#4847)
1e4d227 Introduce an early stop threshold for Kmeans (#4894)
4a0fba1 Add FastScanCodeScanner dispatch boundary with per-SIMD TUs (#4897)
0fcba22 Add sentinels to index format enums. (#4907)
40480fb Support limits on index deserialization loops — useful for tests (#4902)
1e4544a Add defaulted SIMDLevel template parameter to handler and scaler types (#4867)
abdd37b Add Pixi installation option to docs (#5214)

Changed

9d5491a Templatize SIMDResultHandler on SL, mark handle() final (#5223)
b63f236 Upgrade GitHub Actions CI from CUDA 12.6 to CUDA 13.2 (#5207)
a9f5baa Move RISC-V fast_scan forwarders to dedicated impl-riscv.cpp (#5216)
8932716 Validate SVS storage_kind via shared helper at all deserialization read sites (#5204)
bc490b5 Make IVFPQSearchCagraConfig dtype fields settable from Python (#5191)
ffd3727 Update pattern for index factory string with SVSVamana (#5201)
f37041d Extract Metal shaders to standalone .metal file with kernel wrapper class (#5167)
d98364e perf: reserve() to avoid rehashing/reallocation (#5193)
4eee3d7 Add tolerance and assertion log to test_OIVFPQ (#5188)
2322afd faiss: parallelize post-BLAS reduction loop and end_multiple() in result handlers (#5185)
967eda6 Replace FAISS_ASSERT with FAISS_THROW_FMT for unsupported SVSStorageKind (#5182)
715725d Enable verbose mode for TestPyPI publish step (#5181)
5e12311 Filter wheel artifacts before publishing to PyPI (#5179)
28b2b66 Inline PQ code distance kernels into scanner TUs (#5159)
c41405d Enforce minimum training set size for ScalarQuantizer, check nullptr (#5141)
44c1972 Enable ccache in GitHub Actions cmake builds (#5157)
e09de2d Harden compute_query_factors: add preconditions, handle aliasing, remove dead checks
6a9936c Hoist per-iteration vector allocations in IndexIVFRaBitQFastScan::compute_LUT
c3460c9 Convert binary index Hamming callers to dynamic dispatch (#5071)
ecac651 Convert hamming.cpp public API to dynamic dispatch (#5070)
8d1b964 Enable dynamic dispatch build on Windows (#5127)
a25bcb1 Rely on cuVS for default values (#5140)
0622841 Panorama Optimizations (#5041)
f4338ad Parameterize existing tests with @for_all_simd_levels (#5134)
9d56749 Wire 512-bit QBS kernels into fast scan DD dispatch (#5075)
bdf877f facebook-unused-include-check in IndexNNDescent.cpp (#5132)
632427f fbcode/faiss/invlists (#5048)
e23c661 Replace dispatch_HammingComputer with with_HammingComputer (#5126)
ed7e1f2 Hoist SIMD dispatch outside loops in 4 call sites (#5074)
582246b Remove use of __AMDGCN_WAVEFRONT_SIZE macro with it being deprecated in ROCm 7 (#4619)
2dff119 Add braces around single-line control flow statements
cc0c087 Hoist SIMD dispatch out of approx_topk hot loops (#5073)
18f93fd Replace SSE3 intrinsics in Index2Layer with portable scalar code (#5076)
42765ab Optimize RaBitQ FastScan LUT construction for high-dimensional data (#5110)
ec5e70c Optimize multibit sign-bit unpacking in RaBitQ FastScan handlers (#5097)
3f127ee Thread qb/centered search params through FastScan LUT and handler (#5095)
3c19c52 Use std:: qualified math functions to avoid float-to-double promotion (#5088)
48d2b0c Replace typedef with using declarations (#5086)
e12debc Remove extra semicolons after inline member function definitions (#5085)
5679d3a HNSW Cagra base-level entrypoint sampling parallel (#5068)
c4efc42 Remove bare simdlib type aliases from simdlib_dispatch.h (#5069)
4a63c9e Convert partitioning to dynamic dispatch (#5062)
f226be4 Roll out @for_all_simd_levels to key test files (#5061)
afd6949 fbcode/faiss/impl/HNSW.cpp (#5059)
658c434 Replace bare assert() with FAISS_THROW_IF_NOT in core index files (#5052)
364749e Re-enable ROCm runner for AMD GPU CI (#4854)
faeac86 Update outdated C++17 and Python 2 references (#5045)
77e6028 Use HTTPS for GitHub links in README (#5043)
91843bd Use explicit ISA flags instead of -march/-mtune=sapphirerapids (#5034)
acac823 Merge .gitignore updates from faiss-gpu-cu132 into main (#4996)
8d8881e Make distance_compute_blas_threshold dimension-aware (#5022)
553c78b HNSW SIMD dispatch (#5015)
499b488 close the loop on ivf rabitq fast scan filter integration (#5026)
aef066a SVS 0.3.0 (#4999)
cc34603 reuse unpack buffer in multibit IVFRaBitQFastScan refinement (#5003)
d81584d Update Python type stubs: fix class hierarchy and add missing base classes (#5004)
d1c432f Convert rabitq_simd.h to runtime SIMD dispatch (#4912)
6bca961 Convert distances_fused/ to runtime SIMD dispatch (#4911)
04d0d56 facebook-unused-include-check in distances_simd.cpp (

Note truncated.

Open source →

Additional notes2 sources agree

Added

b7618fadc7fdf9b677277fbe96864c1902a6d8dc Add SuperKMeans: faster k-means via ADSampling+PDX progressive pruning (#5168)
66c9d082651659fb6e78d67c547b035d3c2c8228 Add SuperKMeans foundations: math primitives, PDX layout, SIMD kernels
dc7afcfdf0d4bf7cc694f4974db77e471422d4cb Support SVSVamana as IVF coarse quantizer via index_factory (#5175)
7a8e4ddaca3d25363a79942e9fcd26de78436607 CI: cross-compile for riscv64 with RVV dynamic dispatch (#5184)
03202795b6451f2c6942262c998de3b980e96157 Introduce RVV (#5156)
417c53e0ce23e8c3f2843bc727c9493e35b1d5d0 Add NoneSIMDLevel context manager for cross-level reference checks (#5158)
46def46902231ed57552ddbcfd751e17720c6d32 Enable SVS IVF Index in FAISS (#4801)
585ba7954afb4da180a897bf195acd7516a44632 Add pip install support via scikit-build-core + cibuildwheel (#4862)
799bf3aaeca910113fd2f1b0287568f07eb002fa Introduce early stop facilities for IVF, attempt 2 (#5160)
66cea524339ee7cd1b3954e3b5a677edb2adad74 Add Metal GPU backend for Apple Silicon (IndexFlat) (#5144)
17fd3332c77c2801da4b467de67b5cadf045b328 Add database-parallel flat search for few-query workloads (#5000)
c7f07bbc0ad0e9c0b68c6c7a3696850f42500906 Add C++ per-SIMD-level tests for distance utility functions (#5154)
0242758a3d073be0f5b0117b524a1fbdfcf02860 Add bit-exact RQ encode cross-level test (#5155)
2f68a3730777f88da34d562bc90f637bb39c1cb3 Add per-SIMD-level tests for hamming.h utils + fix crosshamming bug (#5153)
a7116ed4ec2ab4f0889d305592f579d5b2fb5745 Add TurboQuant (CPU) (#5049)
11e5d3916054b23bf60253763abf5e9033fca39e Add optional persistent locks to IndexHNSW for incremental adds (#5031)
e82aa28b838d1cad66044c3888b75eb832a39fb4 Add OMP exception capture helpers and migrate IndexIVF (#5111)
6299bff1f60e5423fa8cfe8e8d3926f8d87bdafb Add QT_0bit to ScalarQuantizer for centroid-only IVF distance (#5079)
3030fe0635b743ccfddf3684752438269a5ab0c5 Add filtered search for cuVS indexes (#4858)
f83ea51eaa60987315fcd2deca153924abd7995d Support IVFRaBitQSearchParameters in RaBitQFastScan scanner (#5081)
82a823513d8a7a7e497de11e7da0cba2aae7721d Add iterator callbacks for distance computation and heap admission (#5082)
b068fd985a7460c04ec6c4f909900409392520d0 Support IO_FLAG_MMAP_IFC on Darwin (#5058)
113bf6ceac4e9d1bfeaf29279b29a8293759f83b Expose fast_scan_code_size from non-IVF FastScan indices (#5077)
786d993dbfb8b74b2a0d4d9543b2b08310acabbe Expose a fast scan code size from fast scan indices (#5064)
95f1e45221ede8caf1b0139a3c8864d473e0e47d OMP build diagnostics (#4842)
c4ac852e5b5aa2ba4bde8a977633c367aa85d5db Add a balanced assignment function (#5050)
8402a6b764c024ca78fb87d768cb81de3f60c925 Add unit tests for HNSW deserialization validation (#5033)
1dc5e1fb60405eda268a7d55b1d5a14feef962ba V2 enable specifying a subset of SIMD levels that are implemented at dispatching time. (#4959)
e6f5c0c9a9f70f63c00d3ac5651edce6aebd734d Add SIMD dynamic dispatch migration guide (#4973)
8177a3bca62af1042d955277cafe9db73d5a0c1d Add nvidia-smi diagnostic call (#4960)
67fc664fe0ac1774bd4dcf2ab1aaa488cc7d0545 Add CappedInvertedLists for per-list size limiting (#4880)
e57a8939c91244009c7ef5745d59f33139dc46d2 introduce SingleQueryResultCollectHandler (#4926)
748c0311e423844f25d46a95d27d1621c520a4c7 HNSW: add prune_headroom to avoid O(n^2) pruning/locking, headroom test (#4847)
1e4d2271a6b27c70c4d824cc47067d36998d9daf Introduce an early stop threshold for Kmeans (#4894)
4a0fba1fe2c5e5c8fb3405e614ef9b11524b1064 Add FastScanCodeScanner dispatch boundary with per-SIMD TUs (#4897)
0fcba2284c266f12ead208ef61559c81e2d7ac72 Add sentinels to index format enums. (#4907)
40480fb303fe06a7126e78c71965041911b687c5 Support limits on index deserialization loops — useful for tests (#4902)
1e4544a37e62f89bce0baf3253d214d37fba5192 Add defaulted SIMDLevel template parameter to handler and scaler types (#4867)
abdd37bd964b87be5283334ca49ff603a8e46f4c Add Pixi installation option to docs (#5214)

Changed

9d5491a117b1628dd911d942353156db63553fcf Templatize SIMDResultHandler on SL, mark handle() final (#5223)
b63f236a5252580e1f24158ca7bc6f58517983f2 Upgrade GitHub Actions CI from CUDA 12.6 to CUDA 13.2 (#5207)
a9f5baa58c3bacf208d9f02362ea9db47ffb9f21 Move RISC-V fast_scan forwarders to dedicated impl-riscv.cpp (#5216)
8932716646db11b5f03f1d6e64d2b96ad6cd202b Validate SVS storage_kind via shared helper at all deserialization read sites (#5204)
bc490b56b0896a56d8fc68df42b00798ab5d7eee Make IVFPQSearchCagraConfig dtype fields settable from Python (#5191)
ffd3727d4815a819b3c229adc1ec4fb0e584156c Update pattern for index factory string with SVSVamana (#5201)
f37041d0732bed52e186263466e0cc02fd2b6fe2 Extract Metal shaders to standalone .metal file with kernel wrapper class (#5167)
d98364eb490f0f53b9bece0046b2933f1ec7e646 perf: reserve() to avoid rehashing/reallocation (#5193)
4eee3d7859d4872dc30172092e3ff2a7b919df99 Add tolerance and assertion log to test_OIVFPQ (#5188)
2322afd2d8489bcaae4cb50d105786121335b1a7 faiss: parallelize post-BLAS reduction loop and end_multiple() in result handlers (#5185)
967eda65dec2515ccd18c5abe464b485f8a883d0 Replace FAISS_ASSERT with FAISS_THROW_FMT for unsupported SVSStorageKind (#5182)
715725df362a04d7f1424942583c309ee79c14c6 Enable verbose mode for TestPyPI publish step (#5181)
5e123115f109ae17062c0a5b120a21e7723bbc43 Filter wheel artifacts before publishing to PyPI (#5179)
28b2b66cf279ff2bdb7b8a7fd387c7ea328cb642 Inline PQ code distance kernels into scanner TUs (#5159)
c41405d5253cfd45673728f5544acad07de1eb32 Enforce minimum training set size for ScalarQuantizer, check nullptr (#5141)
44c1972aec4e5d19720263c64dcc5592b06edbdc Enable ccache in GitHub Actions cmake builds (#5157)
e09de2d070ea13ad4fdc65b5c1df97d5455732de Harden compute_query_factors: add preconditions, handle aliasing, remove dead checks
6a9936c60de22ef41138e8b6a8f19df5f180c4ea Hoist per-iteration vector allocations in IndexIVFRaBitQFastScan::compute_LUT
c3460c980be78afac11f6dc1f88b170fd7e70693 Convert binary index Hamming callers to dynamic dispatch (#5071)
ecac6511a271e5441d942054c619dbd34d71e83f Convert hamming.cpp public API to dynamic dispatch (#5070)
8d1b964c11a943f72c6d6f8d7e4e250fd15b5697 Enable dynamic dispatch build on Windows (#5127)
a25bcb1e22e06d4626941bcc0d7b1365bd8b428b Rely on cuVS for default values (#5140)
06228410c6a60c4ca127b00a16e4007d5f0d5353 Panorama Optimizations (#5041)
f4338ade1bfae983ff03164f8ef53921b193f46c Parameterize existing tests with @for_all_simd_levels (#5134)
9d567497ecae53cebafa7c81dc4a75b7d1aff8cd Wire 512-bit QBS kernels into fast scan DD dispatch (#5075)
bdf877f90c955581d74d559a5b85719cee78ba0f facebook-unused-include-check in IndexNNDescent.cpp (#5132)
632427fd0d9d11990f9520d09794203151308253 fbcode/faiss/invlists (#5048)
e23c66175d7d5469aee00c9939ccb1ecf7bb4ac5 Replace dispatch_HammingComputer with with_HammingComputer (#5126)
ed7e1f2435b6c7d76d049473758ef50001547812 Hoist SIMD dispatch outside loops in 4 call sites (#5074)
582246b6bd7ad71bf60d6ddf5b4b1066de24ffd3 Remove use of __AMDGCN_WAVEFRONT_SIZE macro with it being deprecated in ROCm 7 (#4619)
2dff119f138167b915d03fa8df23068802f38e51 Add braces around single-line control flow statements
cc0c08754851a6d95870f259c91271fd4560416d Hoist SIMD dispatch out of approx_topk hot loops (#5073)
18f93fd82f1de86245b4566c53553316671112e5 Replace SSE3 intrinsics in Index2Layer with portable scalar code (#5076)
42765abc0061a00e0afc5f32e53c5d798a3ded61 Optimize RaBitQ FastScan LUT construction for high-dimensional data (#5110)
ec5e70c6555c2f95f131d33c731f72f13065d3af Optimize multibit sign-bit unpacking in RaBitQ FastScan handlers (#5097)
3f127ee3bad4922b1ba0cf2f17f1368d99a241ec Thread qb/centered search params through FastScan LUT and handler (#5095)
3c19c52ca0380bbbd032ba15cd9ff8143e4b7b8e Use std:: qualified math functions to avoid float-to-double promotion (#5088)
48d2b0c946de235f6f5cc2a6fb072f90544a2aee Replace typedef with using declarations (#5086)
e12debcb6123eee4dd49eafd143ac20952ad75b1 Remove extra semicolons after inline member function definitions (#5085)
5679d3ad42533e6f367fdd746b60e3b16fd692c7 HNSW Cagra base-level entrypoint sampling parallel (#5068)
c4efc422192b6e43a7f5b2014f5cdd53e8bd8e84 Remove bare simdlib type aliases from simdlib_dispatch.h (#5069)
4a63c9e415078a79a7e95d52eb8c974d3ddefdaa Convert partitioning to dynamic dispatch (#5062)
f226be432c261df5f24c163434e3b45d3abf4193 Roll out @for_all_simd_levels to key test files (#5061)
afd69493777f8a95c3b2c810f35a24aef8105c8c fbcode/faiss/impl/HNSW.cpp (#5059)
658c4340c0a6ff548c291d26d7182e09e703344a Replace bare assert() with FAISS_THROW_IF_NOT in core index files (#5052)
364749e88624ae2c93c099e1a798c48a47c1ab2d Re-enable ROCm runner for AMD GPU CI (#4854)
faeac8636c4c6c15c47fa60fc7f2c18db775b3b7 Update outdated C++17 and Python 2 references (#5045)
77e602870bd860043db843ccbf458c2ef1c534b2 Use HTTPS for GitHub links in README (#5043)
91843bdbd3eaf1b4e409b1d61763cb2e35170a12 Use explicit ISA flags instead of -march/-mtune=sapphirerapids (#5034)
acac823db4cfff3fdd4e6a033543e0649a9f9f27 Merge .gitignore updates from faiss-gpu-cu132 into main (#4996)
8d8881ee3f6a5a9f8de827742ca38f9db89b78ec Make distance_compute_blas_threshold dimension-aware (#5022)
553c78b25685d0aa38a2844ab35fdd9ed6801e57 HNSW SIMD dispatch (#5015)
499b4885b8aaf439013a820b6f0f14e57ad4ff1a close the loop on ivf rabitq fast scan filter integration (#5026)
aef066ac0ebc914a7de2d6ab1770df15a1b10c69 SVS 0.3.0 (#4999)
cc34603106c16d30edbdd9cb1fcd35ce8f2fbc5a reuse unpack buffer in multibit IVFRaBitQFastScan refinement (#5003)
d81584d548191ff002c304f5d5c8e3f519e6e4e2 Update Python type stubs: fix class hierarchy and add missing base classes (#5004)
d1c432fb5fd7d1bf96a73299e95fa6476b2825e3 Convert rabitq_simd.h to runtime SIMD dispatch (#4912)
6bca9613b7822ee7392957bf19cc53c237133bd4 Convert distances_fused/ to runtime SIMD dispatch (#4911)
04d0d56d0e24ad8da8283a210f8f0527ae2ac91c facebook-unused-include-check in distances_simd.cpp (#5013)
7eb05f01fab7bce7750f84f4ce63fb17354cdd43 facebook-unused-include-check in partition.cpp (#5012)
7e04fa26212bbfba8749f837fc5e6be1360613b4 Update Python type stubs (init.pyi) to match current API (#4998)
db0e798d5a93a97589d7705ae44f0f7b222bea90 Use omp_get_max_threads() for OpenMP thread count (#4991)
9af8384f14edb7c47c8ac6869c6dd45884f72369 Bump openblas from 0.3.30 to 0.3.32 in conda recipes (#4994)
c048917dcd686fcdc7dd386c6d2b345767085bac Enable quieting FAISS warning "inverted lists not stored" (#4964)
b2ddb465bf8fe6a3cf470a67d3f8d70df8cc86c9 Loosen SVS findpackage version requirement (#4969)
d8c3d26130552d70b32c7d61acb9a315f66b045a Upgrade to C++20 for CUVS (#4881)
94f2b51fdbf9d72956e84131a12dc6d911e7c362 Rename train() with queries to train_with_queries (#4955)
02d609caff186bbbf04a2b541ffa1fd7a7a967fc Upgrade cuVS Version to 26.02 (#4945)
55a47c927cd93f22c2b60953bc9e1138e31f1ec4 Upgrade cuVS Version to 26.02, switch various BUCK files over to 26.02 (#4788)
0d147a78bc60574d6ae03d09e9e04c9b32d5b6c5 Make READVECTOR byte limit configurable (#4928)
9962fbec6ee8cd0ea1d8a05c9e8d435652d6a096 HNSW: narrow critical section during add to avoid lock contention (#4915)
b1feeb79f88ea11e2c7cbebb2a9fed7e9aa67bc2 Convert approx_topk and residual_quantizer_encode_steps to SIMDLevel dynamic dispatch (#4923)
58a57e6bca746821b8a018a35bd6d7057565de03 Re-enable backwards compatibility tests for rabitq now that 1.14.1 released (#4924)
5dcef1858d960b22226a47d1b9ee326033e73b63 Refactor binary HNSW stats to use OpenMP reduction instead of destructor sync (#4910)
d8c1a9785f21c84e7517226f6d7f84bfbfd26a36 move distances implementation into SIMD specific compile units (#4906)
ea11f0ef9c04d346a65e9096e13d6d5cb6ba8aab Switch all search paths to FastScanCodeScanner and remove make_knn_handler (#4904)
57190f91cddffc8ac168dd5ac88044976d4be864 Wire RaBitQ search through FastScanCodeScanner dispatch (#4903)
79e8acb07228e2b706ce784358a9bfc42672761f Parallelize compute_residuals in IndexIVFPQ (#4654)
eddea2a6078c362f19deb39fd16ccff875121a9e Extract RaBitQ result handler to impl/fast_scan/ (#4895)
734751a16f6babc42a850b1c78b7cdb09bc9ddbb Extract PQ4 kernels to includable headers in impl/pq_4bit/ (#4868)
bb298a2323c1df5fd4ee4cdf8ddae23d6614e45c convert simdlib in distances_simd.cpp (#4884)
8d8268c8aeea11cb0f25f37789882d4800ac89f1 Templatize simdlib types on SIMDLevel (#4866)
c74809a7c6f0830d1b6090494979533252465269 Improve exception safety for the 'own_fields'/allocation pattern (#4864)
c6cf004ecefebfc71822b72be617ef96839ae004 use level but not 0 in neighbor_range(i, level, &begin, &end) (line 198) (#5005)

Fixed

cb69d7c5665241d1e84f476105ffb6fd6c7c192c Fix Dq=None crash and np.empty nondeterminism in search_preassigned wrappers (#5221)
5c92c5c1ee744545a77f6be9f14d4fdccaad7869 Fix flaky test_hnsw smoke test by increasing efSearch (#5222)
8d3cc92536e340e210f05c8265b8523e1efd1168 Fix clone_index null return for IndexRowwiseMinMax (#5220)
ff1d54397780abc22dd90e958bef11c77ea88a1b Fix python_unnecessary_generator_set_comprehension issues in faiss/tests/test_fastscan_filter.py (#5215)
e2359156e08c96bdc9a6ddb2ed28e97963f8f3c4 Fix FastScan DD regression by threading SIMD level through kernel functions (#5210)
172324aea695fca9f5d59c7a6a7098f28e314e52 Fix IDSelectorBitmap conversion to cuVS bitset (#5211)
6bd749e39bd17d45acfdb6f92d5027fc64ffeced Fix IDSelector leak via SearchParameters.sel setter (#5208)
6376bc35e3db976d55a8fcec039dd1ffb444aca9 Fix CI: (cuVS) conda shards cache lock, (ARM) bump openblas, (SVS) fix LeanVec double-destruction (#5209)
4b5a73577ee90c8466139f5448ef2b5db033808b Fix python_unnecessary_generator_set_comprehension issues in faiss/.../bench_fw/optimize.py (#5205)
6789bf5eb7ff1ffac95ca24b4473af8ba8024ac5 Validate IndexIDMap id_map and inner-index ntotal consistency to prevent search-time null deref (#5203)
f7508155124af8de6ca427b9842feb770d0075f2 Pin MKL to <2026 to avoid soname mismatch with pytorch (#5192)
bb2ce7137d32901898b81fb519abee00c2ce841e Fix 7 broken tests (#5197)
f323c0fa11a3e17b4417d81fb75159304b5a550f Fix backward compat test (#5195)
23cd94cd51062f014cb48a1b1a00884311ad786e Reduce BinaryCagra test parameters to prevent CI timeouts (#5194)
85fc627c1e7559557e1d0c427eb506e124944524 Validate ProductQuantizer M*ksub during deserialization to prevent oversized allocations (#5187)
6cef1bb08c47ff7fc772e8d70a931b014eab771c Fix cuVS and ROCm CI conda environment failures (#5180)
4a471eebf9e41059ad4f677172bdda6a0af504a2 Pin conda <25.7 to fix Windows build (#5176)
a9c0d4187b508565599677e4421143a676d3625c Work around GCC 12 miscompilation of AVX2 histogram (#5124)
40a8cc003460855a810b2063191eaac0382bb85e Fix lints associated with early stop facilities for IVF (#5172)
1e69d5ab20223212e23df9630a5240b60d075f04 Fix peak-memory spike when loading IVF invlists via IO_FLAG_MMAP_IFC (#5122)
9942229393c6a161cff4061634805d2e5c24adf6 Fix add_sa_codes silently accepting non-int64 ids, corrupting stored labels (#5171)
c5ddbc710427d6b323f0ee0bdfa406f77c42d63e c_api: fix IndexShards own_indices getter/setter name mismatch (#5165)
8380e2543db6ae43feba7fd210207bd916eca953 Fix avx512 unit test (#5161)
71448c07d309a5058703a81e1dda6059c4a9f947 Fix issue with svs tests (#5162)
6c70444d0843d6de7e1e8376c0cc4526cae1025a Validate code_size during deserialization to prevent oversized allocations (#5151)
9dbb81ca9d674b031090c4ac31cf001a65fdd78f Set own_fields after reading Index2Layer sub-quantizer during deserialization (#5147)
01b22b4cc94fe4121de30f04b54a1f451a1e6188 Fix memory leak warning for unordered_multimap in IndexIVFFlatDedup Python binding (#1667) (#5145)
130fc2481a0fa768848387cd3f724395157cebe1 Fix flaky test (#5150)
dfce6e948e153b783d8bd0785c1c30432e84a166 Fix OMP exception safety in IndexHNSW search (#5133)
c6273349f948dc2df83be83c8f49c12ca63842e8 Fix sa_decode offset bug and integer overflow in IndexRefine (#5143)
1b4b995b97c97e503c051e7b3087b2937aed5633 Fix implicit integer precision loss from 64-bit to 32-bit (#5091)
b0568942efcb980f91fb286e1fdebfa0925f1150 Fix miscellaneous lint warnings (#5093)
ed11f283d63915c014d1e176033a1bd0f1fd72ff Fix broken fbcode//faiss/tests:test_contrib - test_checkpoint (test_contrib.TestBigBatchSearch) (#5139)
277c53d299ac7e059d2791d2cb15956819287586 Fix PCA training bug and memory safety issues in VectorTransform.cpp (#5138)
83125407637fa69358de00ffcf7681947109d92a Fix integer type truncation in IDSelectorBatch bloom filter mask (#5136)
9cbc8da28efb865d9f8fed2f20ccf80ca2db0d20 Fix integer overflow and unbounded loop in Clustering.cpp (#5130)
3c4056d822129f2e211d9aa8fc1ee1436fed82a9 Fix race condition in HNSW::add_with_locks (#5129)
6e64c5d6873e9ba7a66918fe6f743997178b665b Bump GCC pin in faiss-gpu conda recipe to fix AVX2 SIMD miscompilation (#5125)
58f6ebb7fa84a995a3c11c7b4e8d3a9a0adaa489 Add per-read byte limit to SVS ReaderStreambuf to prevent OOM from corrupt index data (#5118)
817ecf9cd444dbd7461d3a09a0035b16448d2c6f Validate inverted list entry sizes against deserialization byte limit (#5117)
e6be16241a10e1143dff263150315c220cc2a100 Validate VectorTransform dimension consistency during deserialization (#5115)
6707eacfaffc012bdb2ac95eae37b4baceef11f0 Add is_trained check to IndexIVF search and range_search to prevent querying untrained indexes (#5114)
349df7035bd2ddfa30f148c2e6225e90be91cdfa Validate IndexHNSW2Level storage type during deserialization and search (#5113)
edd6f3b54c38e3ca063368f5079d2662b33098a1 Reject null quantizer during IVF index deserialization (#5112)
27078c52809b8ad43fbc73b6857a0865fe9c59c0 Fix OMP exception safety in IndexNNDescent and IndexNSG search (#5106)
99db159afd3f5e85f55dd959ef4cdb62e15086ad Fix OMP exception safety in IndexFlatCodes search (#5105)
40ad64694d4096b606a72a72622647a9421e8789 Fix cuVS build (#5107)
57bf47410f58616d1a104523327d474cdc264fb1 Fix int/size_t signedness mismatches in HNSW add (#5116)
f9f116d303e93abd5bf79bc89d4227505e086448 Validate k_factor during deserialization to prevent search-time OOM (#5104)
2d8232c981f4148719bedbe9d0dcc2cd98f5fa3a Set code_size on IVFRaBitQFastScanScanner (#5099)
b923091c0d87c68d2ff5e7076c5c1b1cf6489aed fix: GPU CAGRA copyFrom host-memory lifetime dependency and add regression test (#4968)
796977a1abc1ebfc42ce27fda655593089bcc70e Fix IndexBinaryMultiHash::reset() not clearing hash maps and avoid unnecessary copies (#5100)
b86066a9989ea7223826e06d25170b9e03c67b66 fix: cagra bug (#4963)
91f6636b2a4d08c13c06e99dd0a1cb0bc77ceaf9 Add value-initialization to uninitialized member variables (#5089)
61ed7dc917b2db31edef74e9f825f1d04efbe80e Fix orphaned OpenMP directives in IndexRowwiseMinMax training functions (#5096)
740d3a396575114156cb2cd5e36ba98669a9998d Fix FastScan for indices without own inverted lists + performance optimizations in RaBitQ handler (#5080)
7cc63422d701ee9178ae0f7c8afa8aa6b9533a83 Suppress unused static constexpr variable warnings (#5087)
aa3ce376ced6741b170ceef26bebbbf1f139cbec Fix race condition, memory management, debug output, and hashtable lookup in sorting.cpp (#5078)
86d33710702b668c1413d3bafe7be6048718f138 Validate FastScan M2 consistency during deserialization (#5056)
67f066f7a02f76d3178baccf4c31b4839ff0fee8 Validate SVS storage_kind during deserialization (#5055)
e899476ceb8e21d1739851e2d7d6c0c70cac43e5 Default-initialize SVSStorageKind in IndexSVSVamana (#5054)
c1e48a2ce12a008cd741f4d340af8026807cd020 Add search-path bounds checking to IndexIVF::search1 (#5040)
97d257433a59d7150a52491439f80314f2474cfb Fix correctness bugs in NNDescent Nhood copy/move operations and gen_random bounds (#5072)
06ef481afa0910b07b33eee47d52c542e06a988a Fix uncaught exceptions in IndexIVF::range_search_preassigned OpenMP region (#5053)
255250985a74e81a9a81822959588339dfae0cd0 Fix bugs in NSG search_on_graph and sync_prune (#5063)
7047dc5346abcfcf0a71b58f29aa9c138093af68 Fix priority_queue constructor inheritance for Apple Clang on macOS (#5065)
e0a1aec811bb78af64c140fe47720cf9a80431cc Assorted Dynamic Dispatch cleanup fixes (#5060)
25296bad4f726aacff414a1f467f8a5f6eb0e306 Fix type mismatch bug and remove dead code in IndexRowwiseMinMax::sa_decode_impl (#5051)
7b2705c2698354ca75732970ba2e3526dc8330c7 Debug manifold crash with NaN (#5025)
9d3f97e620a24543f5e4d6e921473de0d63f2cd9 Fix duplicate words and grammar typos across codebase (#5044)
84262f09d4d0bc83c7147ddb53a97b402c2e0539 Fix d_out check bug and add descriptive error messages to VectorTransform assertions (#5047)
ef28c6beb081d265cb6bdace8a3e85606bfd8a56 Fix uncaught exceptions in IndexIVF::search_preassigned OpenMP region (#5037)
8008ca2a70d7806694be80e365ff4b242982b699 Fix deadly signal in to_svs_metric() for unsupported metrics (#5032)
4b8da6820ead7ddd07f7d54e55bb4eb8bb2b8417 Fix IndexSVSVamana null-deref in deserialize_impl (#5029)
ab3da522a4da1cd370052e36106f9743428d06b7 Cap ZnSphereCodecRec decode cache size (#5028)
d68050d269c0aeb1bf7fe0ef5f5f662dba006477 Add missing input validation to IndexFlat::range_search and compute_distance_subset (#5038)
c6fa50a4510760be0075e9cd0383b2290dec5057 Validate IndexIDMap id_map size during deserialization (#5024)
408786bec9d016f2b9893373a2ad40b82868eeb4 Validate AdditiveQuantizer dimension consistency during deserialization (#5023)
0ce4dfb2069019f9390fbfbf50be3052443325d5 Fix AVX512_SPR build failure for 512-bit SIMD types (#5030)
a308adae08ae7b589e2f11da1ec08fa0d46b93c1 Harden ScalarQuantizer deserialization validation (#5020)
ab46dcbeaed4817617f214604bdffc40026b5922 Validate qbs at deserialization time in FastScan index formats (#5019)
668da4a38684b95083e1ce737483d3ca64279c24 Validate IndexBinaryHash b <= code_size*8 during deserialization (#5011)
87bc9b08722d3a5ec01a91f5b16236fc56caf3bc Validate IndexFastScan quantizer state during deserialization (#5009)
4c391564a60e9e4441d5aeb3744347d855530e79 Validate codebooks size at deserialization time in read_AdditiveQuantizer (#5008)
8c1f3f48d48147777a0a2adaf3ceb661e0c43a7d read-heap-buffer-overflow (size 32) in float faiss::fvec_norm_L2sqr<>()
226ccf3e84d4439070262ff72dadab7de1d0e5f7 Fix OSX arm64 nightly: bump libopenblas from 0.3.30 to 0.3.32 (#5006)
aeb170c15a105d640118bd32c37e9de81c7432be large-malloc (15023964028 bytes) in __sanitizer::RunMallocHooks() (#5018)
8550f2304af244a4a3ccbd8edb0f2d00aea11320 Validate VectorTransform dimensions during deserialization (#5007)
833d4aaf7b719f30a022271bab7b0d0917edfbff out-of-memory in __gnu_cxx::new_allocator<>::allocate() (#5014)
d716be39ed1092d6e5545809b944b1742283920e Fix avx512_result_handlers build on aarch64 (#4965)
18c85e17e1ed5605c814d284e2704bca612f0e70 Add deserialization vector limit enforcement for ResidualCoarseQuantizer (#4997)
b74f3cd8d9ba57a0549e91154057b8554c0fdd21 Validate IndexBinaryMultiHash b and nhash*b bounds during deserialization (#4984)
b9f49601d109ae4c88adfec1ff84b5f40c00a2f3 Add invlists null check to IndexBinaryIVF::search (#4980)
ba1981ca4ce1c6139e80a6d00ad6c26e44443dec Add input validation to IndexBinaryHNSW search preconditions (#4979)
d26c15d954a1d0f372a4da59ee30622d314a666d Validate binary index consistency during deserialization (#4978)
2c28ef5382a21ba623afe56e3fdade97ae93b50c Validate RaBitQ qb during deserialization (#4983)
5a4d62146c606796b24711003bf18d87402f2400 deadly-signal (vector::_M_default_append) in __clang_call_terminate (#4986)
0759009bd7ddf290170c930c70a722badd05aa7d Fix compiler warnings for pip wheel builds (macOS + Windows) (#4989)
8a80b3dd6e2bcd1f0c6bef98aa396ac6d5a53e0d signal in faiss::ArrayInvertedListsPanorama::ArrayInvertedListsPanorama() (#4987)
12675affbcf1976c77a75d16813135f92584debc out-of-memory in __gnu_cxx::new_allocator<>::allocate() (#4985)
0a8e516058cb76e5e026858dc3ded478aa815683 Fix HNSW Panorama Perf Bug (#4974)
6dbf2713db8673d8007514d6f1d6375c72d5c99f Validate VectorTransform data during deserialization (#4981)
3f7938924d611b38adfdb00b7da27f24531decb7 Fix MSVC link failure: remove __restrict from fvec_madd specializations (#4972)
d47b281578d98c25e3187201e6f138ba91f50e97 deadly-signal (vector::_M_default_append) in __clang_call_terminate (#4982)
f851c54cd9fb89bda2680bbdeb28298d10c8178e large-malloc (6241124352 bytes) in __sanitizer::RunMallocHooks() (#4977)
bd5ed00c384d5513917c51e348694d899e01718f Validate inverted lists pointer in IndexIVF operations (#4951)
398857b1ff7248579ab8a63ee5d570d0bfcce71e Handle empty index in IndexHNSWCagra base_level_only search (#4950)
d647b596e01e1df4b79d5c513410b33463619aa8 Validate graph index data during deserialization (#4949)
e5114c2368073d00316a866baa33e6595df3c396 fix: resolve compiler warnings in binary and miscellaneous Index files (#4940)
44f57ba4c91d2e4933f99f2c102df3ef5a3994cf fix: resolve compiler warnings in top-level FAISS source files (#4943)
b3a2914982f5affcc61f280bfe3860b97c686f1f fix: resolve compiler warnings in test files (#4944)
88efcb981a22fabd9a6dbe6cd583c51362864420 fix: resolve compiler warnings in core base classes and headers (#4933)
7acc420929c40abc44899417b6e49e2eb28bb06e fix: resolve compiler warnings in inverted lists and cppcontrib (#4939)
64a2122ef1c8142287f277ec9e00cd2f8132d80a fix: resolve compiler warnings in IO, result handlers, and misc impl (#4936)
a05f13ff6a395ad6916b6772ef4baa0d8ba72d26 fix: resolve compiler warnings in PQ, PreTransform, and Refine Index files (#4941)
7645426ac338b61856210a9d8a5eee9d653d47e4 fix: resolve compiler warnings in IVF Index implementations (#4942)
f5142bfc5d8a8deed0268336c95352bab71dd781 large-malloc (7046432072 bytes) in __sanitizer::RunMallocHooks() (#4967)
4db01e3431b88b7363c55406b0babd6e8812977f fix: resolve compiler warnings in graph-based index implementations (#4956)
ccc3cd619a8e2bc6ee970e61b1f4f7417724aadc Validate code vector sizes during index deserialization (#4948)
ec30fb387c79a99454805d07f47311e77fd90bfc Use BitstringReader in IndexBinaryHash to avoid OOB reads (#4961)
fa8ddd99a323ac289a96fd37b38e8c40f34004ad Fix flaky test for Panaorama/regular IVFFlat equivalence (#4954)
1e6729d53e5080e8290d1b7993969c3267f4c393 Validate quantizer data sizes during index deserialization (#4947)
62a19690988c0765e95d6707bb563a2233a9d2aa fix: resolve compiler warnings in distance and hamming utilities (#4937)
645a742b51813f11b033f9ebda909f0e60771dd5 Re add cuvs to cmakelists.txt (#4953)
5287db053721511ffa15a563d22665ea4ad02ae4 fix: resolve compiler warnings in graph-based index implementations (#4934)
a48e45af3a7405a502711a1c63aeef5ba9b35ad1 fix: resolve compiler warnings in quantizer implementations (#4935)
32ca42f873053080435e57e306c3db123b3ea901 fix: resolve compiler warnings in SIMD, sorting, and misc utilities (#4938)
52d9fc705391db885138e86df42dbaf97d7f0a55 Add overflow check in READVECTOR macro (#4946)
0f7e65e53de6dd4a0273a3791fa70075b9de77f4 Enforce memory limit in read_ProductQuantizer (#4930)
9ea026cc93f349eeca25b1246af1e97745df84f7 Validate codes vector size in BlockInvertedLists deserialization (#4920)
a4385dfd079f3de6046f6d68d56322600efdbaad Validate n_per_block and block_size in BlockInvertedLists deserialization (#4919)
cdb725466736f1667d2f2a9426594e639097a967 Validate id_map size matches ntotal in IndexBinaryIDMap deserialization (#4917)
05ca04e469a6458f9abe25073f034987cb9844e4 Validate per-entry ilsz in read_binary_multi_hash_map (#4916)
b023178c5dfc8980dd7dee40ef228198b13ef65a Fix MSVC OpenMP build: use idx_t instead of size_t in parallel for loop (#4922)
5b83ec68569b82ece3ca182f45b6951c761f8aff Fix OpenMP critical section contention in IndexBinaryHNSW search (#4909)
e96ba2d13cb827a19cc3c6fe4996e8f963a07d79 fix: decouple coarse quantizer from cuvs index reset (#4885)
9d6b2e7b186cdc889c91866aff2ca2b5907ffb4d Additional input validation for index deserialization (#4899)
796bdf94a8878ebef1725fcea40f9dab06d1c5e5 Additional binary index input validation (#4898)
d0434be649f556fbbefb7317899805a48f077553 Fix cloning for IVFFlatPanorama (#4887)
3e4c103d77a7b2cd1759f359fe073385920571f5 Validate dynamic_casts during index load (#4883)
47e53b4468a87fe4c33609b0dfd1902328e3ea6f Fix backward compatibility to use latest version (#4855)

Deprecated

812010fe19e1cac03e5f0bf529ddae9cb0ea4f1b Remove flaky test (#5169)
74dceee9038339bb57e232cfd655df3cdde3891c Remove RaBitQStats debug counters from all RaBitQ handlers (#5102)
000eec283a89342c411820de35bf9b118f39602c Delete old pq4_accumulate_loop files (replaced by dispatching.h) (#4905)

1.13.017 Nov 2025Deprecated

Deprecated 2705d7a Delete rocm runner for now until it is fixed ( #4658 ) f9ccd58 Remove invalid assertion checking #neighbors == graph degree

Full Changelog: v1.12.0...v1.13.0

Added
2cf82ca Implement PanoramaStats (#4628)
859127c Implement serialization for IndexIVFFlatPanorama (#4636)
7744239 Add getter/setter for balanced_bins in PCAMatrix C API issue #4617 (#4630)
e6510bd Add immediate notification when autoclose label is applied (#4624)
f983e3a Integrate Panorama into IndexIVFFlatPanorama (#4606)
3af3e00 Implementation of IndexIVFRaBitQFastScan (#4596)
f58fd4c Add InvertedListScanner for IVFPQFastScan (#4537)
01d394e RaBitQ Fast Scan (#4595)
752832c Add missing Thrust includes (#4597)
6470b8d Simplify RaBitQ slightly, improving speed and recall (#4550)
bc3e3a1 RaBitQ: SIMD helper simplification, use faster popcount for doc-side sum (#4573)
b7c88ea Microbenchmarks for rabitq_simd.h (#4572)
786e405 RabitQ test coverage for SIMD codepaths (#4571)
fa55327 Adding Idx tags in extended APIs (#4532)
fbbb290 Set NN Descent Metric From CAGRA Params (#4540)
514b44f OSS changes: Enable ROCm to work with Faiss on BUCK (#4485)

Changed
a9cc039 Upgrade cuVS to 25.10 and build pkg with CUDA=12.6 (#4639)
6ca45d2 Optimize IVFRaBitQFastScan query factors: nnlist to nnprobe (#4643)
49df773 Update GitHub workflow to use clang-format-21 (#4644)
a0bd7aa clang-format | Format fbsource with clang-format 21.
595c8aa Allow unaligned fast-scan (#4623)
4fab13c Upgrade Faiss OSS side to numpy2 (#4523)
2505168 Expose Remaining IVF-PQ params for CAGRA (#4593)
1ed2611 Remove unused imports from faiss directory files (#4565)
e5de66e Use Development.Module component in CMake FindPython (#4549)
3c7235c Update Install Docs with Correct cuVS Version (#4547)
f361df8 Use c++11 only in headers (#4421)
dd637c9 Change extended API suffix from Ex to _ex (#4530)

Fixed
d94c330 chore: fix typos in some files (#4669)
5d9f8d4 Fix additional typos found in second review (#4667)
451ca84 Fix typos in documentation, tutorials, and C API (#4666)
d0d066e Fix typos in comments and documentation (#4664)
3ffec12 Fix typos in comments and documentation (#4662)
2af54a4 Fix typos in comments and documentation (#4661)
ec877e7 Fix typos in comments and documentation (#4660)
3de200f try to fix nightly (#4657)
18f5574 typo: SIFT -> GIST (#4642)
675661b Fix IndexIVFRaBitQFastScan nprobe handling in search_with_parameters (#4629)
64b1f3a Fix autoclose workflow - add GH_REPO for immediate notification
e5fee10 Fix IVFFlatPanorama bench (#4622)
70df32b Fix IndexIVFRaBitQFastScan by overriding search_preassigned (#4618)
61c1c76 Fix nightly build on windows due to c++20 initializer (#4612)
513eabc Replace static constexpr with inline constexpr in header (#4613)
97dc014 fix: initializing order of gpu compile options (#4581)
dbc7550 Fix ARM64 compilation error in IndexRaBitQFastScan (#4611)
484dd97 Add IndexBinaryIDMap2 support to index binary factory (#4603)
041ac84 Fix memory bloat in IndexBinaryHash search due to argument copying in dispatch_HammingComputer (#4600)
2964a37 Fix nightly by updating mkl version (#4604)
7f7b518 Unable to import faiss in python in AIX (#4602)
266b712 Added code to catch 'ModuleNotFoundError' exception (#4577)
3b14dad Fix AIX compilation issue while building python extension (#4587)
1deba7b Add attribute validation to prevent silent failures in SWIG wrappers (#4583)
3671c61 more formatting (#4568)
d98ff43 Fix FAISS build with ROCm7 (#4567)
50b3eb4 add "override" to overriden destructors (#4566)
2135e5a add missing explicit specifier (#4564)
0031d61 Revert D80734790 (#4563)
ee6b7dd add #pragma once to some header files (#4562)
8b83ebb fix python blank space lints (#4560)
f46ac53 Fix missing object reference in Faiss python wrapper for IndexIVFRaBitQ (#4554)
69f1ac0 resolve Open Source requirement violations (#4551)
dca887a Remove unused include from IVFlib.cpp (#4534)
3b3bf5e bugfix: add a macro guard for avx512 operation (#4539)
fe5e77f Remove unnecessary std::move on temporaries to fix -Wpessimizing-move warnings in DeviceTensor-inl.cuh (#4545)
2fca92f fix test_binary_cagra tests failure (#4546)
8d9d3be Remove unused includes from AutoTune.cpp (#4533)
75be84d Remove unused imports from fbcode/faiss (#4543)
5c61ed8 Fix E302 lint errors: Add required blank lines before class definitions (#4541)
daceaac replace type(x) == y with isinstance(x, y) (#4542)

Deprecated
2705d7a Delete rocm runner for now until it is fixed (#4658)
f9ccd58 Remove invalid assertion checking #neighbors == graph degree (#4528)

Open source →

Additional notes2 sources agree

Added

2cf82cabf2b2150ca76b9949377b484f109a94d1 Implement PanoramaStats (#4628)
859127cd0b5e9f2ee52e095211202b0f5dbb6414 Implement serialization for IndexIVFFlatPanorama (#4636)
7744239dc59c0cc9c665949f0d533daac40e567a Add getter/setter for balanced_bins in PCAMatrix C API issue #4617 (#4630)
e6510bd00478d563e380604374f41a27e8697ce6 Add immediate notification when autoclose label is applied (#4624)
f983e3ab69c46ecb7728395613ace22729bbd4a8 Integrate Panorama into IndexIVFFlatPanorama (#4606)
3af3e00103079554b1071d2dcea7ccedc9693a44 Implementation of IndexIVFRaBitQFastScan (#4596)
f58fd4c8ca0fdf6c230c062ed70b93586ae9ec28 Add InvertedListScanner for IVFPQFastScan (#4537)
01d394e5837f10c6270c7284c9f5e306c5cd87c4 RaBitQ Fast Scan (#4595)
752832cac9c4038f52916a0ed2a839401671ce00 Add missing Thrust includes (#4597)
6470b8d9d0f9c0adc71df6d5a1ce64199be85305 Simplify RaBitQ slightly, improving speed and recall (#4550)
bc3e3a1ef1464f1b90c6aa4ba4f075582e4f539a RaBitQ: SIMD helper simplification, use faster popcount for doc-side sum (#4573)
b7c88eaff82766c69cd3ab7a3a57fa4d4efb603f Microbenchmarks for rabitq_simd.h (#4572)
786e4051fc7940825e8f6f5247bc269f4bc8bd11 RabitQ test coverage for SIMD codepaths (#4571)
fa5532734e40dbf66b571634ff597fe314ace054 Adding Idx tags in extended APIs (#4532)
fbbb290a5762e265a501324b3bfeb04c1cb5b006 Set NN Descent Metric From CAGRA Params (#4540)
514b44fca8542bafe8640adcbf1cccce1900f74c OSS changes: Enable ROCm to work with Faiss on BUCK (#4485)

Changed

a9cc039aed084bfb824a84eb7fc749ff0e84370c Upgrade cuVS to 25.10 and build pkg with CUDA=12.6 (#4639)
6ca45d23a7c10761d380daa6b4029c347f30e4ef Optimize IVFRaBitQFastScan query factors: nnlist to nnprobe (#4643)
49df7737f868eeb75df450718fc1155d93c43d35 Update GitHub workflow to use clang-format-21 (#4644)
a0bd7aac694cd43e3ba0a371cbd7bacf31ded564 clang-format | Format fbsource with clang-format 21.
595c8aa8a23815abd3682a2c08b0b359b37eb621 Allow unaligned fast-scan (#4623)
4fab13c9c67b5402343ca722c83ff7a65a9a48ba Upgrade Faiss OSS side to numpy2 (#4523)
2505168e1870318126aee2ac2cb416b1fe55376f Expose Remaining IVF-PQ params for CAGRA (#4593)
1ed2611c032fe2c509b0f781ee8eee3760142e52 Remove unused imports from faiss directory files (#4565)
e5de66e2c24c319e73df22858b520fcfa269d300 Use Development.Module component in CMake FindPython (#4549)
3c7235c6c7040c1c09e39095cc60d794f73b4a35 Update Install Docs with Correct cuVS Version (#4547)
f361df8385ccbf9c47b85a6f8e45c71f8f656283 Use c++11 only in headers (#4421)
dd637c98d60f51b96c9d2457ebfa319e3f881a47 Change extended API suffix from Ex to _ex (#4530)

Fixed

ff1a2d4f6eaf4216c74e6006bf55791cf135b767 Fix typos in core library (root + impl/) (#4670)
d94c33065a240e460603352770ec6aa2dc70b8cf chore: fix typos in some files (#4669)
5d9f8d484cffd0961c374c1c7253c8c406b889df Fix additional typos found in second review (#4667)
451ca8429d39f82ed6a933ac874068fa5f13252c Fix typos in documentation, tutorials, and C API (#4666)
d0d066e430ec0711121926cadaaeb85fd3cbad1f Fix typos in comments and documentation (#4664)
3ffec120fe411495939dc070e8588afb1db1b7b1 Fix typos in comments and documentation (#4662)
2af54a46bb0dd3a5ba84388c0830f44067bf6e53 Fix typos in comments and documentation (#4661)
ec877e794e26cc0778e3665349060e3c224dcb8c Fix typos in comments and documentation (#4660)
3de200f5afb0452b934458c1b235860b4f2089ee try to fix nightly (#4657)
18f5574a50416ab9b61d4806cc802166663912b7 typo: SIFT -> GIST (#4642)
675661b4d363b0eba314d38624324b16b548883f Fix IndexIVFRaBitQFastScan nprobe handling in search_with_parameters (#4629)
64b1f3a04cf730c78f43bb071fcc6afa1b48a992 Fix autoclose workflow - add GH_REPO for immediate notification
e5fee1015a90dd3667c0d1f7d866bbc2c023c475 Fix IVFFlatPanorama bench (#4622)
70df32b2e989e0c47c04cfd78be3221c79c8d408 Fix IndexIVFRaBitQFastScan by overriding search_preassigned (#4618)
61c1c76929809914240b9407162f54513cd2b7db Fix nightly build on windows due to c++20 initializer (#4612)
513eabc91d52a9ebc11545ba0b21f3fd1ca99b32 Replace static constexpr with inline constexpr in header (#4613)
97dc014472480ff23dd0bf77c5aa46c073d59687 fix: initializing order of gpu compile options (#4581)
dbc75506150e923abdaf4d09baf4fac56a20740e Fix ARM64 compilation error in IndexRaBitQFastScan (#4611)
484dd97ce07ed56129660cccba1d634f765e73ea Add IndexBinaryIDMap2 support to index binary factory (#4603)
041ac84318dac299739e568fafee318317ede6ef Fix memory bloat in IndexBinaryHash search due to argument copying in dispatch_HammingComputer (#4600)
2964a374dbe5d9359cb0f794f73cc334d95c3ac4 Fix nightly by updating mkl version (#4604)
7f7b518df5f1a21fd4591ed68222c26532580467 Unable to import faiss in python in AIX (#4602)
266b71285aabfe3ca66715b7878b1d8fd1e472f7 Added code to catch 'ModuleNotFoundError' exception (#4577)
3b14dad6d9ac48a0764e5ba01a45bca1d7d738ee Fix AIX compilation issue while building python extension (#4587)
1deba7b90f21d952c86affe79721a06ec5800907 Add attribute validation to prevent silent failures in SWIG wrappers (#4583)
3671c61af1455d50c169e5ddd5d6a18d81174c3a more formatting (#4568)
d98ff432b98a4d7bea770c0cc221d8a0dab2e4a9 Fix FAISS build with ROCm7 (#4567)
50b3eb48d26b32a2beedb570d1fcf7757830bfcb add "override" to overriden destructors (#4566)
2135e5a28d0f2a932dd02b64b566ac1a52266540 add missing explicit specifier (#4564)
0031d61da81dab33579f7481f6773549ca816fa8 Revert D80734790 (#4563)
ee6b7ddd9df954492b597b49b63e8760923317b7 add #pragma once to some header files (#4562)
8b83ebbf6271b713264671fcd128c33def70e8d3 fix python blank space lints (#4560)
f46ac530b534f98b8cd5b78373e9ccabad0fdee4 Fix missing object reference in Faiss python wrapper for IndexIVFRaBitQ (#4554)
69f1ac0f0fa635812fdba5180074981f5d7ba2de resolve Open Source requirement violations (#4551)
dca887a656e943352bcf1cae6dbaaf6771ac272e Remove unused include from IVFlib.cpp (#4534)
3b3bf5e4e53eb0846fd1109432773a25a879b01b bugfix: add a macro guard for avx512 operation (#4539)
fe5e77f1cb0bd26a5a86d9fc1e99f0b2ecacab93 Remove unnecessary std::move on temporaries to fix -Wpessimizing-move warnings in DeviceTensor-inl.cuh (#4545)
2fca92f477063a1188a1cf52d3414b3eba05936b fix test_binary_cagra tests failure (#4546)
8d9d3bea7c2d9879e6ce4919d1a25954ef533c8f Remove unused includes from AutoTune.cpp (#4533)
75be84d086190cfda8a516a5f276b5ba0e8cc706 Remove unused imports from fbcode/faiss (#4543)
5c61ed8c94296d0ca97cecd53cbcef91f63e2de4 Fix E302 lint errors: Add required blank lines before class definitions (#4541)
daceaac99aeb9dfc8577a0dc6f47c3f6f9b39055 replace type(x) == y with isinstance(x, y) (#4542)

Deprecated

2705d7a5d30f121ffc7553d2d7087fce1a448d6b Delete rocm runner for now until it is fixed (#4658)
f9ccd582f9a9b8400428625d1ff3217ae83422b9 Remove invalid assertion checking #neighbors == graph degree (#4528)

faiss-cpu

Release timeline

Releases

[1.15.0] - 2026-07-31

Highlights

Full Changelog

Highlights

Full Changelog

[1.13.2] - 2025-12-19