microsoft/onnxruntime-genai v0.13.0

v0.13.0

Repository: microsoft/onnxruntime-genai

Tag: v0.13.0

Published: 2026-04-15T19:56:04Z

Prerelease: no

Release notes:

What's Changed

• update WebGPU buffer memory info name by @fs-eire in https://github.com/microsoft/onnxruntime-genai/pull/1957
• Add enable_profiling in Runtime Options by @xiaofeihan1 in https://github.com/microsoft/onnxruntime-genai/pull/1949
• Fix uninitialized tools variable and improve exception debug messages by @sheller-ms in https://github.com/microsoft/onnxruntime-genai/pull/1971
• Add common download to Phi-3 tutorial by @kunal-vaishnavi in https://github.com/microsoft/onnxruntime-genai/pull/1973
• Add support for InternLM2 model architecture by @amdrajeevp1 in https://github.com/microsoft/onnxruntime-genai/pull/1958
• Update cmake cuda architecture and use win-arm64 pool workaround by @baijumeswani in https://github.com/microsoft/onnxruntime-genai/pull/1976
• Update examples after 0.12.0 release by @kunal-vaishnavi in https://github.com/microsoft/onnxruntime-genai/pull/1980
• Add CI pipeline for WebGPU EP model testing by @qjia7 in https://github.com/microsoft/onnxruntime-genai/pull/1956
• Fix Python nightly build by @kunal-vaishnavi in https://github.com/microsoft/onnxruntime-genai/pull/1981
• Add missing Quark 0.11 weight patterns for ChatGLM3 output layer by @poganesh in https://github.com/microsoft/onnxruntime-genai/pull/1983
• Support Qwen2.5-VL pre-quantized models in qwen.py by @poganesh in https://github.com/microsoft/onnxruntime-genai/pull/1985
• [VitisAI] external_ep_libray support fix for WinML by @akholodnamdcom in https://github.com/microsoft/onnxruntime-genai/pull/1984
• Fix guidance bug by @baijumeswani in https://github.com/microsoft/onnxruntime-genai/pull/1988
• Fix incorrect batch responses when using multiple prompts by @lnigam in https://github.com/microsoft/onnxruntime-genai/pull/1986
• Enable webgpu graph capture in base.py by @qjia7 in https://github.com/microsoft/onnxruntime-genai/pull/1991
• Harden CUDA error checking across the codebase by @Copilot in https://github.com/microsoft/onnxruntime-genai/pull/1994
• allow pruned models for prefill by @fs-eire in https://github.com/microsoft/onnxruntime-genai/pull/1995
• Fix WinML Packaging Pipeline by @baijumeswani in https://github.com/microsoft/onnxruntime-genai/pull/1998
• Add small changes after pruning prefill by @kunal-vaishnavi in https://github.com/microsoft/onnxruntime-genai/pull/2000
• webgpu: Optimize Copyfrom by @qjia7 in https://github.com/microsoft/onnxruntime-genai/pull/1992
• Add support for CUDA 13 by @baijumeswani in https://github.com/microsoft/onnxruntime-genai/pull/2001
• add webgpu to qmoe path by @guschmue in https://github.com/microsoft/onnxruntime-genai/pull/2005
• Fix ERNIE 4.5 model builder: rope_attrs and config architecture name by @xiaoyao9184 in https://github.com/microsoft/onnxruntime-genai/pull/2007
• Bug fix in Continuous Decoding by @chilukam-qti in https://github.com/microsoft/onnxruntime-genai/pull/2008
• Update Phi-4 mm README links by @kunal-vaishnavi in https://github.com/microsoft/onnxruntime-genai/pull/2014
• Add Qwen3-VL model support + multi-image input support in Qwen VL family by @hanbitmyths in https://github.com/microsoft/onnxruntime-genai/pull/2003
• Add Qwen3.5 model support and optimize multi-image handling by @apsonawane in https://github.com/microsoft/onnxruntime-genai/pull/2019
• Reuse a single generator via RewindTo(0) in benchmark instead of creating multiple generators by @qjia7 in https://github.com/microsoft/onnxruntime-genai/pull/2002
• [RyzenAI] WinML compatibility fix by @akholodnamdcom in https://github.com/microsoft/onnxruntime-genai/pull/2026
• Nemotron ASR Support for Streaming by @nenad1002 in https://github.com/microsoft/onnxruntime-genai/pull/1997
• [WebGPU] Fix the prefill regression when graph capture is ON by @qjia7 in https://github.com/microsoft/onnxruntime-genai/pull/2021
• Support 4 inputs for nemotron model by @jiafatom in https://github.com/microsoft/onnxruntime-genai/pull/2036
• Updated java packaging based on python packaging logic by @EPNW-Eric in https://github.com/microsoft/onnxruntime-genai/pull/2029
• Fix android packaging pipeline by @baijumeswani in https://github.com/microsoft/onnxruntime-genai/pull/2039
• Add OpenAI's Whisper to model builder by @kunal-vaishnavi in https://github.com/microsoft/onnxruntime-genai/pull/2018
• [Java] Add a dependency on onnxruntime (#2030) by @EPNW-Eric in https://github.com/microsoft/onnxruntime-genai/pull/2040
• Fix mutually exclusive inputs for language models by @kunal-vaishnavi in https://github.com/microsoft/onnxruntime-genai/pull/2046
• Decouple plugin execution providers (EPs) from the USE_WINML pre-processor macro by @baijumeswani in https://github.com/microsoft/onnxruntime-genai/pull/2038
• Route pipeline model RunOptions through SetRunOption for proper special key handling by @Copilot in https://github.com/microsoft/onnxruntime-genai/pull/2044
• Add ort_build_version and ort_build_source parameters to nuget and python packaging pipelines, remove ROCm support by @Copilot in https://github.com/microsoft/onnxruntime-genai/pull/2049
• Add batched multi-image vision path and window_size config for Qwen VL by @hanbitmyths in https://github.com/microsoft/onnxruntime-genai/pull/2050
• docs: fix formatting and syntax highlighting in documentation by @riddles-the-one in https://github.com/microsoft/onnxruntime-genai/pull/2051
• Add Silero VAD Support to Nemotron Streaming ASR by @sayanshaw24 in https://github.com/microsoft/onnxruntime-genai/pull/2035
• Add Qwen3.5 hybrid decoder export support (GatedDeltaNet + Attention) by @apsonawane in https://github.com/microsoft/onnxruntime-genai/pull/2043
• Add support for QNN stateful models by @qti-ashimaj in https://github.com/microsoft/onnxruntime-genai/pull/2012
• Allocate recurrent state via device allocator to enable CUDA graph capture by @apsonawane in https://github.com/microsoft/onnxruntime-genai/pull/2057
• Speed up CI pipelines by @Copilot in https://github.com/microsoft/onnxruntime-genai/pull/2052
• Fix tool calling for TRT-RTX models by @kunal-vaishnavi in https://github.com/microsoft/onnxruntime-genai/pull/2048
• Fix vision pipeline EP hardcoding and pixel_values rank mismatch for Qwen VL models by @apsonawane in https://github.com/microsoft/onnxruntime-genai/pull/2060

New Contributors

• @sheller-ms made their first contribution in https://github.com/microsoft/onnxruntime-genai/pull/1971

*…