NewYour coding agent can read the release notes before it upgrades.Set up the MCP server →
PyPI
PyPI
MLX-VLM is a package for inference and fine-tuning of Vision Language Models (VLMs) and Omni Models (VLMs with audio and video support) on your Mac using MLX.
Latest stable
0.7.12
Last release 18 days ago
04 Nov 2026
Ships fairly regularly
a new release about every 2 weeks
Nearly every release is documented
notes for 60 of the last 60 stable releases
Nothing withdrawn
no release was ever pulled
3 years old
93 releases · first in 2024
Drag across the ruler to compare, or tap a release to open it.
Lower MOSS-Image peak memory during VAE decode by @felix-ab in https://github.com/Neywa-Labs/mlx-vlm/pull/2615
Full Changelog: https://github.com/Neywa-Labs/mlx-vlm/compare/v0.7.11...v0.7.12
Your coding agent can read these notes before it upgrades. Set up the MCP server →
feat(server): live-input STT WebSocket on the audio worker by @TheAngryPit in https://github.com/Blaizzy/mlx-vlm/pull/2167
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.7.10...v0.7.11
Add Muse Spark Open support with audio/video/images by @lucasnewman in https://github.com/Blaizzy/mlx-vlm/pull/2542
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.7.9...v0.7.10
Fix MiniCPM vision dtype with quantized embeddings by @shiftedx in https://github.com/Blaizzy/mlx-vlm/pull/2511
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.7.8...v0.7.9
adding kolibri1 support by @Lazarus-931 in https://github.com/Blaizzy/mlx-vlm/pull/2423
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.7.6...v0.7.8
Stream compaction progress and recover oversized requests by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/2418
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.7.5...v0.7.6
Enable Qwen3-Omni multimodal batching by @Lazarus-931 in https://github.com/Blaizzy/mlx-vlm/pull/2349
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.7.4...v0.7.5
Generate multiple images per prompt via num_images (ming_image) by @Lazarus-931 in https://github.com/Blaizzy/mlx-vlm/pull/2353
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.7.3...v0.7.4
Hemmingway-1 support by @Lazarus-931 in https://github.com/Blaizzy/mlx-vlm/pull/2330
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.7.2...v0.7.3
Downmix stereo audio before resampling in load_audio by @kiarina in https://github.com/Blaizzy/mlx-vlm/pull/2258
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.7.1...v0.7.2
Add standalone DINOv2 model with register tokens and mask support by @eklipse2k8 in https://github.com/Blaizzy/mlx-vlm/pull/2184
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.7.0...v0.7.1
Optimize Qwen4 n-gram shard gathers by @pyros-projects in https://github.com/Blaizzy/mlx-vlm/pull/2034
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.6.17...v0.7.0
Optimize Qwen4 n-gram shard gathers by @pyros-projects in https://github.com/Blaizzy/mlx-vlm/pull/2034
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.6.17...v0.7.0rc0
Fix MiniMax-M3 sparse-index compile re-trace leak (Metal 499k resource limit) by @ziomancer in https://github.com/Blaizzy/mlx-vlm/pull/2016
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.6.16...v0.6.17
Warn when resize_shape is ignored by a custom image processor by @eptan in https://github.com/Blaizzy/mlx-vlm/pull/1829
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.6.15...v0.6.16
Stop a batched row's answer depending on its neighbour's length by @Lazarus-931 in https://github.com/Blaizzy/mlx-vlm/pull/1946
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.6.14...v0.6.15
Fix multimodal server prefill alignment by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/1872
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.6.13...v0.6.14
Preserve APC prefix reuse for growing prepared prompts by @byoan in https://github.com/Blaizzy/mlx-vlm/pull/1713
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.6.12...v0.6.13
Fix LFM tool call parsing and bump version to 0.6.11 by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/1792
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.6.10...v0.6.12
Add and improve MLX-VLM agent skills by @Lazarus-931 in https://github.com/Blaizzy/mlx-vlm/pull/1747
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.6.9...v0.6.10
Add Kimi K3 by @kernelpool in https://github.com/Blaizzy/mlx-vlm/pull/1746
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.6.8...v0.6.9
Vendor standard models by @Lazarus-931 in https://github.com/Blaizzy/mlx-vlm/pull/1702
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.6.7...v0.6.8
Support Laguna S by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/1650
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.6.6...v0.6.7
vendor Gemma/Phi/GLM/Cohere text models and reuse shared MLPs by @Lazarus-931 in https://github.com/Blaizzy/mlx-vlm/pull/1616
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.6.5...v0.6.6
Add deepseek_v3 and deepseek_v32 model types by @Lazarus-931 in https://github.com/Blaizzy/mlx-vlm/pull/1517
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.6.4...v0.6.5
Add support for TTS and STT endpoints in the server by @lucasnewman in https://github.com/Blaizzy/mlx-vlm/pull/1358
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.6.3...v0.6.4
Fix APC splitting on multimodal inputs by @lucasnewman in https://github.com/Blaizzy/mlx-vlm/pull/1311
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.6.2...v0.6.3
Make sure to evaluate cached MRoPE parameters on model load by @lucasnewman in https://github.com/Blaizzy/mlx-vlm/pull/1272
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.6.1...v0.6.2
Move linear spec handling to model backends by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/1259
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.6.0...v0.6.1
fix: replace deprecated use_fast=True with backend=torchvision in load_processor by @Jonathangadeaharder in https://github.com/Blaizzy/mlx-vlm/pull/11…
finish_reason inconsistencies by @spicyneuron in https://github.com/Blaizzy/mlx-vlm/pull/1215usage and timings for server endpoints by @spicyneuron in https://github.com/Blaizzy/mlx-vlm/pull/1216not use_rht (fix masked decode under RHT) by @popfido in https://github.com/Blaizzy/mlx-vlm/pull/1244Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.5.0...v0.6.0
Fix gemma4 multi-image processing for different-sized images by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/938
json_schema response_format support by @avbiswas in https://github.com/Blaizzy/mlx-vlm/pull/1047--max-tokens to server by @spicyneuron in https://github.com/Blaizzy/mlx-vlm/pull/1120Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.4.4...v0.5.0
Fix Gemma 4 chunked prefill for KV-shared models and thinking by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/901
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.4.3...v0.4.4
Add SAM 3.1 with Object Multiplex and optimized realtime pipeline by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/880
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.4.2...v0.4.3
Ensure minimal metadata exists in model card by @pcuenca in https://github.com/Blaizzy/mlx-vlm/pull/853
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.4.1...v0.4.2
feat(server): add --model and --adapter-path flags for startup preloading by @auggie246 in https://github.com/Blaizzy/mlx-vlm/pull/811
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.4.0...v0.4.1
Fix gemma3n short prompts by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/751
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.3.12...v0.4.0
[MODEL] support qwen3.5 series by @JJJYmmm in https://github.com/Blaizzy/mlx-vlm/pull/722
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.3.11...v0.3.12
Refactor input embedding handling in _generate_batch function by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/694
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.3.10...v0.3.11
Fix qk_norm for lighton by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/615
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.3.9...v0.3.10
Fix qwen3_vl ValueError: Image features and image tokens do not match by @ziya32 in https://github.com/Blaizzy/mlx-vlm/pull/608
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.3.8...v0.3.9
Add chat_ui command by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/589
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.3.7...v0.3.8
update readme with new openai endpoints details by @mguella in https://github.com/Blaizzy/mlx-vlm/pull/585
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.3.6...v0.3.7
Fix: Input cast error by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/559
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.3.5...v0.3.6
Remove docs actions temporarly by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/536
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.3.4...v0.3.5
Add n_kv_heads property used in LM Studio to glm4v_moe.LanguageModel by @hehua2008 in https://github.com/Blaizzy/mlx-vlm/pull/472
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.3.3...v0.3.4
fix changelog task by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/441
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.3.2...v0.3.3
Fix energy calc in omni.py by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/427
pyproject.toml by @SauravMaheshkar in https://github.com/Blaizzy/mlx-vlm/pull/282Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.3.1...v0.3.2
fix(chat-ui): Fix imports, blocking Chat-ui server start by @zenyr in https://github.com/Blaizzy/mlx-vlm/pull/420
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.3.0...v0.3.1
[gemma3n] Correctly scale text embeddings for quantized gemma3n conversions by @neilmehta24 in https://github.com/Blaizzy/mlx-vlm/pull/397
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.2.0...v0.3.0
Fix Gemma 3 config by @DePasqualeOrg in https://github.com/Blaizzy/mlx-vlm/pull/384
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.1.27...v0.2.0
Fix README POST request typo by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/358
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.1.26...v0.1.27
When converting models, default dtype to config.json's torch_dtype by @neilmehta24 in https://github.com/Blaizzy/mlx-vlm/pull/323
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.1.25...v0.1.26
Fix snapshot download by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/318
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.1.24...v0.1.25
Deps: Remove Scipy MLX by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/301
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.1.23...v0.1.24
Add internVL3 by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/295
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.1.22...v0.1.23
chore: fix sliding mask by @FL33TW00D in https://github.com/Blaizzy/mlx-vlm/pull/272
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.1.21...v0.1.22
fix remaining DEFAULT_TEMP by @omercelik in https://github.com/Blaizzy/mlx-vlm/pull/265
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.1.20...v0.1.21
Fix return tensors by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/262
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.1.19...v0.1.20
Copy all json files by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/249
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.1.18...v0.1.19
Fix Gemma 3 Text-only and prompt template by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/248
Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.1.17...v0.1.18
One column per month.