# vLLM v0.8.5 — Release & Upgrade Intelligence
- **Product:** vLLM (https://vllm.ai)
- **Version:** v0.8.5
- **Released:** Apr 28, 2025
- **Impact Score:** 7/10
- **Breaking Changes:** Source flagged
- **Canonical Analysis:** https://changelogs.us/vllm/updates/v0.8.5/

## Executive Summary (TL;DR)
vLLM v0.8.5. Source highlight: This release features important multi-modal bug fixes, day 0 support for Qwen3, and xgrammar's structure tag feature for tool calling.

## Breaking Changes & Deprecations
- Breaking changes 🚨 * `--enable-chunked-prefill`, `--multi-step-stream-outputs`, `--disable-chunked-mm-input` can no longer explicitly be set to `False`. Instead, add `no-` to the start of the argument (i.e. `--enable-chunked-prefill` and `--no-enable-chunked-prefill`) (https://github.com/vllm-project/vllm/pull/16533)

## New Features
- Add ModernBERT (#16648)
- Add Granite Speech Support (#16246)
- Add Kimi-VL model support (#16387)
- Add Qwen2.5-Omni model support (thinker only) (#15130)
- Add `structural_tag` support using xgrammar (#17085)
- Validate urls object for multimodal content parts (#16990)

## Improvements
- FA3 decode perf improvement - single mma warp group support for head dim 128 (#16864)
- Update to lastest FA3 code (#13111)
- Support Cutlass MLA for Blackwell GPUs (#16032)
- Support Microsoft Runtime Kernel Lib for our Low Precision Computation - BitBLAS (#6036)
- Optimize rotary_emb implementation to use Triton operator for improved performance (#16457)
- Improve configs - `SchedulerConfig`

## Bug Fixes
- Day 0 support for Qwen3 and Qwen3MoE. This release fixes fp8 weight loading (#17318) and adds tuned MoE configs (#17328).
- Accuracy fixes for Llama4 Int4 (#16801), chat template for Llama 4 models (#16428), enhanced AMD support (#16674, #16847)
- Use safe serialization and fix zmq setup for mooncake pipe (#17192)
- [fix]: Dockerfile.ppc64le fixes for opencv-python and hf-xet
- [Bugfix] Multi-modal caches not acting like LRU caches
- [TPU][V1] Fix exponential padding when `max-num-batched-tokens` is not a power of 2

## Upgrade Command
```bash
python -m pip install 'vllm==0.8.5'
```