●Change8

Migrating to vLLM v0.9.1

Version v0.9.1 introduces 6 breaking changes. This guide details how to update your code.

Released: 6/10/2025

6
Breaking Changes
5
Migration Steps
9
Affected Symbols

⚠️ Check Your Code

If you use any of these symbols, you need to read this guide:

LLMAsyncLLMEngine.generateQwen2EmbeddingModelChatCompletionRequestKVEventBatchget_dummy_textget_dummy_mm_dataasync_timeoutSampler

Breaking Changes

●Issue #1

Positional arguments other than 'model' are no longer allowed when initializing the LLM class; use keyword arguments instead.

●Issue #2

The 'inputs' argument fallback in Engine classes has been removed.

●Issue #3

Fallbacks for the Embeddings API have been removed.

●Issue #4

Default mean pooling for Qwen2EmbeddingModel has been removed; pooling must now be explicitly defined.

●Issue #5

Custom model implementations must now explicitly override 'get_dummy_text' and 'get_dummy_mm_data'.

●Issue #6

Metrics deprecated in version 0.8 have been completely removed.

Migration Steps

  1. 1
    Update LLM class initializations to use keyword arguments for all parameters except 'model'.
  2. 2
    Update custom model classes to implement 'get_dummy_text' and 'get_dummy_mm_data'.
  3. 3
    Review monitoring systems for removed metrics that were deprecated in 0.8.
  4. 4
    Explicitly configure pooling for Qwen2EmbeddingModel if relying on previous defaults.
  5. 5
    Update Engine class calls to remove the 'inputs' argument if still in use.

Release Summary

This release introduces significant performance optimizations for large-scale serving, including DP/EP CUDA graph support and Blackwell hardware integration. It also enforces stricter API usage by removing several long-standing deprecations and positional argument support in the LLM class.

Need More Details?

View the full release notes and all changes for vLLM v0.9.1.

View Full Changelog