# vLLM v0.6.1 — Release & Upgrade Intelligence
- **Product:** vLLM (https://vllm.ai)
- **Version:** v0.6.1
- **Released:** Sep 11, 2024
- **Impact Score:** 5/10
- **Breaking Changes:** Not verified
- **Canonical Analysis:** https://changelogs.us/vllm/updates/v0.6.1/

## Executive Summary (TL;DR)
vLLM v0.6.1. Source highlight: Added support for Pixtral (`mistralai/Pixtral-12B-2409`). (#8377, #8168)

## New Features
- Added support for Pixtral (`mistralai/Pixtral-12B-2409`). (#8377, #8168)
- Added support for Llava-Next-Video (#7559), Qwen-VL (#8029), Qwen2-VL (#7905)
- Add progress reporting to batch runner (#8060)
- Add support for NVIDIA ModelOpt static scaling checkpoints. (#6112)
- Added CODE_OF_CONDUCT.md (#8161)
- [Doc] Add multi-image input example and update supported models

## Improvements
- Memory optimization for awq_gemm and awq_dequantize, 2x throughput (#8248)
- Update the docker image to use Python 3.12 for small performance bump. (#8133)

## Bug Fixes
- [bugfix] Upgrade minimum OpenAI version
- [Core/Bugfix] Add query dtype as per FlashInfer API requirements.
- [BugFix] Fix Granite model configuration
- [Bugfix] Fix Hermes tool call chat template bug
- [Model] Multi-input support for LLaVA and fix embedding inputs for multi-image models
- [tpu][misc] fix typo

## Upgrade Command
```bash
python -m pip install 'vllm==0.6.1'
```