PyPI · #4929 most downloaded on PyPI

mlx-vlm

0.7.1814K downloads/moBlaizzy/mlx-vlm ↗New releases in your daily digest, changelog included.

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.

Last release 3 days ago
14 Sep 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Nothing withdrawn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
2 years old
83 releases · first in 2024

Release timeline

83 releases since 2024

202420252026

Releases

0.1.730 Dec 2024Breaking
This release introduces some breaking changes. If you encounter any issues, please open an issue or submit a PR.
What's Changed

Fix multi-image and 2x speed improvements (DS-VL2) by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/157

Refactor utils (model loading, inference and output processing) by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/161

Fix Llama-3.2-Vision (18x faster generation and 75% less memory usage) by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/163

⚠️ Breaking Changes

This release introduces some breaking changes. If you encounter any issues, please open an issue or submit a PR.

Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.1.6...v0.1.7
Open source →
0.1.622 Dec 2024
Fixes DeepSeek quant loading by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/156
What's Changed

Fixes DeepSeek quant loading by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/156

Fixes Florence-2 LM only by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/156

Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.1.5...v0.1.6

One column per month.

0.1.522 Dec 2024

Add support for Deepseek-vl2 by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/153

What's Changed

Add support for Deepseek-vl2 by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/153
Add support for Language only inputs by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/153

Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.1.4...v0.1.5

Open source →

0.1.45 Dec 2024

Add support for Paligemma-2 by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/142

What's Changed

Add support for Paligemma-2 by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/142
Update prompt utils + bump version (Paligemma-2) by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/143

Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.1.3...v0.1.4

Open source →

0.1.328 Nov 2024

Add lazy eval during conversion by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/127

What's Changed

Add lazy eval during conversion by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/127
Open tokenizer.json within context manager by @neilmehta24 in https://github.com/Blaizzy/mlx-vlm/pull/129
Fix Bugs in chat UI by @terhechte in https://github.com/Blaizzy/mlx-vlm/pull/96
Fix broken stream generate for SmolVLM and others by @andimarafioti in https://github.com/Blaizzy/mlx-vlm/pull/132
Fix idefics3 by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/133

New Contributors

@neilmehta24 made their first contribution in https://github.com/Blaizzy/mlx-vlm/pull/129
@terhechte made their first contribution in https://github.com/Blaizzy/mlx-vlm/pull/96
@andimarafioti made their first contribution in https://github.com/Blaizzy/mlx-vlm/pull/132

Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.1.2...v0.1.3

Open source →

0.1.226 Nov 2024

Fix padding type casting by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/125

What's Changed

Fix padding type casting by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/125
Idefics 3 support by @pcuenca in https://github.com/Blaizzy/mlx-vlm/pull/124

New Contributors

@pcuenca made their first contribution in https://github.com/Blaizzy/mlx-vlm/pull/124

Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.1.1...v0.1.2

Open source →

0.1.123 Nov 2024

Add example notebooks and support for system role by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/95

What's Changed

Add example notebooks and support for system role by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/95
fix pixtral image prompt order for doc VQA by @ndurner in https://github.com/Blaizzy/mlx-vlm/pull/99
Fix Qwen2-VL OCR and repetition penalty by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/109
Qwen2-VL performance improvements by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/113
Faster / more memory efficient Qwen VL by @awni in https://github.com/Blaizzy/mlx-vlm/pull/114
Add support for Molmo by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/112
Add support for Florence-2 by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/105
Fix image masks and update pointing example by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/117

New Contributors

@ndurner made their first contribution in https://github.com/Blaizzy/mlx-vlm/pull/99
@awni made their first contribution in https://github.com/Blaizzy/mlx-vlm/pull/114

Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.1.0...v0.1.1

Open source →

0.1.018 Oct 2024

Add support for Pixtral-12B by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/67

What's Changed

Add support for Pixtral-12B by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/67
Fix pixtral multi-image by @hiima234 in https://github.com/Blaizzy/mlx-vlm/pull/41
Added: Qwen2-VL Unit Tests, Refactored Weight Sanitization by @benzimring in https://github.com/Blaizzy/mlx-vlm/pull/63
Trainer + Multi image v0.1.0 by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/41
Fix example scripts in the readme.md to import and use load_config by @mark-lord in https://github.com/Blaizzy/mlx-vlm/pull/82
Qwen2-VL Improvements (1-2x speedup) by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/89
Fix Paligemma object detection and segmentation by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/90
Add support for Llama-3.2-vision & Resize image by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/83
Fix idefics-2 mask by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/91

New Contributors

@benzimring made their first contribution in https://github.com/Blaizzy/mlx-vlm/pull/63
@mark-lord made their first contribution in https://github.com/Blaizzy/mlx-vlm/pull/82

Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.0.15...v0.1.0

Open source →

0.0.1529 Sept 2024

Qwen2-VL fix vision tower bug for HD imagges by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/62

What's Changed

Qwen2-VL fix vision tower bug for HD imagges by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/62

Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.0.14...v0.0.15

Open source →

0.0.1428 Sept 2024

Add support for Qwen2-VL by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/59

What's Changed

Add support for Qwen2-VL by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/59

Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.0.13...v0.0.14

Open source →

0.0.1316 Aug 2024

fix chat template bug by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/55

What's Changed

fix chat template bug by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/55

Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.0.12...v0.0.13

Open source →

0.0.122 Aug 2024

Refactor KVCache and add stream generate by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/53

What's Changed

Refactor KVCache and add stream generate by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/53
Fix gradio app generation by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/54

Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.0.11...v0.0.12

Open source →

0.0.114 Jul 2024

Add Dolphin-vision and Bunny by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/50

What's Changed

Add Dolphin-vision and Bunny by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/50

Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.0.10...v0.0.11

Open source →

0.0.1024 Jun 2024

Add support for phi-3-vision-128k-instruct by @JosefAlbers in https://github.com/Blaizzy/mlx-vlm/pull/36

What's Changed

Add support for phi-3-vision-128k-instruct by @JosefAlbers in https://github.com/Blaizzy/mlx-vlm/pull/36

New Contributors

@JosefAlbers made their first contribution in https://github.com/Blaizzy/mlx-vlm/pull/36

Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.0.9...v0.0.10

Open source →

0.0.922 Jun 2024

Add support for Llava-Next (v1.6) by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/43

What's Changed

Add support for Llava-Next (v1.6) by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/43

Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.0.8...v0.0.9

Open source →

0.0.88 Jun 2024

Update load_image function to handle BytesIO input by @gabewillen in https://github.com/Blaizzy/mlx-vlm/pull/34

What's Changed

Update load_image function to handle BytesIO input by @gabewillen in https://github.com/Blaizzy/mlx-vlm/pull/34
Add support for DeepSeek-VL by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/37

New Contributors

@gabewillen made their first contribution in https://github.com/Blaizzy/mlx-vlm/pull/34

Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.0.7...v0.0.8

Open source →

0.0.725 May 2024

Fix idefics2 OCR by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/31

What's Changed

Fix idefics2 OCR by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/31

Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.0.6...v0.0.7

Open source →

0.0.624 May 2024

Add support for PaliGemma and Quant Siglip by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/24

What's Changed

Add support for PaliGemma and Quant Siglip by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/24
Remove chat history until multi-image support by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/22
Improve ChatUI by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/29

Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.0.5...v0.0.6

Open source →

0.0.57 May 2024

Better error message for non safetensor repository by @IamShubhamGupto in https://github.com/Blaizzy/mlx-vlm/pull/15

What's Changed

Better error message for non safetensor repository by @IamShubhamGupto in https://github.com/Blaizzy/mlx-vlm/pull/15
Add chat UI with gradio by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/17
fix llava template by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/19

New Contributors

@IamShubhamGupto made their first contribution in https://github.com/Blaizzy/mlx-vlm/pull/15

Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.0.4...v0.0.5

Open source →

0.0.43 May 2024

Update image processor by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/7

What's Changed

Update image processor by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/7
Add tests by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/13
Simpler generate API by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/10
Add support for idefics2 by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/10

Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.0.3...v0.0.4

Open source →

0.0.323 Apr 2024

Fix hf-to-mlx conversion and quantization by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/6

What's Changed

Fix hf-to-mlx conversion and quantization by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/6

Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.0.2...v0.0.3

Open source →

0.0.223 Apr 2024

Update docs by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/4

What's Changed

Update docs by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/4

Full Changelog: https://github.com/Blaizzy/mlx-vlm/compare/v0.0.1...v0.0.2

Open source →

0.0.123 Apr 2024

API redesign for v0.0.1 by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/2

What's Changed

API redesign for v0.0.1 by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/2
Create python-publish.yml by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/1
fix build by @Blaizzy in https://github.com/Blaizzy/mlx-vlm/pull/3

New Contributors

@Blaizzy made their first contribution in https://github.com/Blaizzy/mlx-vlm/pull/2

Full Changelog: https://github.com/Blaizzy/mlx-vlm/commits/v0.0.1

Open source →

Your coding agent can read these notes before it upgrades. Set up the MCP server →

mlx-vlm

Release timeline

Releases

What's Changed

⚠️ Breaking Changes

What's Changed

What's Changed

What's Changed

What's Changed

New Contributors

What's Changed

New Contributors

What's Changed

New Contributors

What's Changed

New Contributors

What's Changed

What's Changed

What's Changed

What's Changed

What's Changed

What's Changed

New Contributors

What's Changed

What's Changed

New Contributors

What's Changed

What's Changed

What's Changed

New Contributors

What's Changed

What's Changed

What's Changed

What's Changed

New Contributors