Starstruk Logo MarkStarstruk

v1.9.0

xorbitsai/inferencev1.9.0Aug 16, 2025by XprobeBot

AI Summary

This release introduces Qwen-Image model support, GLM 4.5 series support, and enhanced rerank model engine switching capabilities.

Key Highlights

  • Added Qwen-Image model support
  • Added GLM 4.5 series support including GLM-4.5v
  • Enhanced rerank model engine switching capabilities

New Features

  • Qwen-Image model support
  • GLM 4.5 series support
  • GLM-4.5v support
  • Enhanced rerank model engine switching
  • gpt-oss model support
  • Function calling support for deepseek-r1-0528

Full Release Notes

# What's new in 1.9.0 (2025-08-16)
These are the changes in inference v1.9.0.
## New features
* FEAT: [UI] running models data display replica. by @yiboyasss in https://github.com/xorbitsai/inference/pull/3897
* FEAT: [model] Qwen-Image by @qinxuye in https://github.com/xorbitsai/inference/pull/3916
* FEAT: [model] gpt-oss by @qinxuye in https://github.com/xorbitsai/inference/pull/3924
* FEAT: function calling support for deepseek-r1-0528 by @qinxuye in https://github.com/xorbitsai/inference/pull/3931
* FEAT: Support for GLM 4.5 quantized models by @Jun-Howie in https://github.com/xorbitsai/inference/pull/3945
* FEAT: sglang support streaming function call by @aniya105 in https://github.com/xorbitsai/inference/pull/3939
* FEAT: parsing harmony format for gpt-oss by @qinxuye in https://github.com/xorbitsai/inference/pull/3948
* FEAT: Add support for switching rerank model engines and support for rerank of vllm engine by @zhcn000000 in https://github.com/xorbitsai/inference/pull/3881
* FEAT: Support GLM-4.5v by @Jun-Howie in https://github.com/xorbitsai/inference/pull/3957
## Enhancements
* ENH: Add qwen3 new model to tool call list by @zhcn000000 in https://github.com/xorbitsai/inference/pull/3900
* ENH: Update chat_template for Qwen3-Coder by @Jun-Howie in https://github.com/xorbitsai/inference/pull/3944
* ENH: add flash_attention control params attn_implementation by @amumu96 in https://github.com/xorbitsai/inference/pull/3951
* ENH: support qwen-image gguf by @qinxuye in https://github.com/xorbitsai/inference/pull/3954
* ENH: clean embedding model cache when using vllm engine by @amumu96 in https://github.com/xorbitsai/inference/pull/3956
* BLD: Downgrade flash-attn to version 2.7.4 by @zwt-1234 in https://github.com/xorbitsai/inference/pull/3953
* BLD: Add Openfst source by @zwt-1234 in https://github.com/xorbitsai/inference/pull/3959
## Bug fixes
* BUG: limit datasets version by @qinxuye in https://github.com/xorbitsai/inference/pull/3943
## Documentation
* DOC: add doc about cu128 docker by @qinxuye in https://github.com/xorbitsai/inference/pull/3899
* DOC: Update xllamacpp doc by @codingl2k1 in https://github.com/xorbitsai/inference/pull/3862
## Others
* Replace @torch.no_grad() with @torch.inference_mode() in Qwen3-Reranker by @yasu-oh in https://github.com/xorbitsai/inference/pull/3911


**Full Changelog**: https://github.com/xorbitsai/inference/compare/v1.8.1...v1.9.0