Starstruk Logo MarkStarstruk

v1.9.1

xorbitsai/inferencev1.9.1Aug 30, 2025by XprobeBot

AI Summary

This release adds support for Qwen-Image model with image editing capabilities, Wan 2.2 model, and various audio and embedding enhancements.

Key Highlights

  • Added Qwen-Image model with image2image and inpainting capabilities
  • Added Wan 2.2 model support
  • Enhanced CosyVoice2 with streaming and non-streaming support

New Features

  • Qwen-Image model with image2image and inpainting
  • Wan 2.2 model support
  • Enhanced CosyVoice2 with streaming and non-streaming
  • qwen-image-lightning support
  • DeepSeek-V3.1 quantization and tools support
  • Seed-OSS-36B model support

Full Release Notes

# What's new in 1.9.1 (2025-08-30)
These are the changes in inference v1.9.1.
## New features
* FEAT: Qwen-Image-Edit by @qinxuye in https://github.com/xorbitsai/inference/pull/3989
* FEAT: Wan 2.2 by @qinxuye in https://github.com/xorbitsai/inference/pull/3996
* FEAT: Update CosyVoice2 to support both streaming and non-streaming speech generation by @Gmgge in https://github.com/xorbitsai/inference/pull/3994
* FEAT: support qwen-image-lightning by @qinxuye in https://github.com/xorbitsai/inference/pull/3995
* FEAT: [UI] support gpu_count configuration in image model. by @yiboyasss in https://github.com/xorbitsai/inference/pull/4016
* FEAT: image2image and inpainting for qwen-image by @qinxuye in https://github.com/xorbitsai/inference/pull/4014
* FEAT: Support Custom vllm embedding dim by @zhcn000000 in https://github.com/xorbitsai/inference/pull/4000
* FEAT: [embedding] support `dimensions` for embedding by @llyycchhee in https://github.com/xorbitsai/inference/pull/3965
* FEAT: [Model] Support DeepSeek-V3.1 Quantization and tool by @Jun-Howie in https://github.com/xorbitsai/inference/pull/4022
* FEAT: Seed-OSS-36B by @Jun-Howie in https://github.com/xorbitsai/inference/pull/4020
## Enhancements
* ENH: added zero shot and voice cloning ability for audio models by @qianduoduo0904 in https://github.com/xorbitsai/inference/pull/3968
* ENH: Add Template for Qwen3 Reranker when model_engine = vllm by @zhcn000000 in https://github.com/xorbitsai/inference/pull/3983
* ENH: Update the environment dependencies for cosyvoice2 by @Gmgge in https://github.com/xorbitsai/inference/pull/4015
* ENH: Compat with xllamacpp 0.2.0 by @codingl2k1 in https://github.com/xorbitsai/inference/pull/4004
* ENH: support chat_template_kwargs for llama.cpp by @qinxuye in https://github.com/xorbitsai/inference/pull/3988
* BLD: Clean up Docker's last legacy cache and images before executing each step by @zwt-1234 in https://github.com/xorbitsai/inference/pull/3963
* BLD: fix CI failures by @qinxuye in https://github.com/xorbitsai/inference/pull/4002
## Bug fixes
* BUG: disable flash_attention when GPU compute capability < 8.0 by @amumu96 in https://github.com/xorbitsai/inference/pull/3973
* BUG: fix rerank model creation by @qinxuye in https://github.com/xorbitsai/inference/pull/3977
## Documentation
* DOC: update models by @qinxuye in https://github.com/xorbitsai/inference/pull/3958
* DOC: add setting limitation of images for multi modal doc by @amumu96 in https://github.com/xorbitsai/inference/pull/4003
* DOC: Update docs about custom models by @OliverBryant in https://github.com/xorbitsai/inference/pull/4019
* DOC: update models & README by @qinxuye in https://github.com/xorbitsai/inference/pull/4023
## Others
* FEAT：KAT-V1 by @Jun-Howie in https://github.com/xorbitsai/inference/pull/3998

## New Contributors
* @qianduoduo0904 made their first contribution in https://github.com/xorbitsai/inference/pull/3968
* @OliverBryant made their first contribution in https://github.com/xorbitsai/inference/pull/4019

**Full Changelog**: https://github.com/xorbitsai/inference/compare/v1.9.0...v1.9.1