Repository Issue Activity (beta)

#853 Exposing a compatible interface for Chrome Built-in AI (Prompt API) / Polyfill

#838 [Bug] Service worker does not send response back to page even when there is keepAlive and heartbeat constantly sent

#843 structural_tag response_format hangs generation with numeric unhandled rejection (0.2.84)

#833 [Bug] Examples can render user/model text as HTML

#844 [Bug] shapeCache LRU eviction disposes in-use ShapeTuples → "Object has already been disposed" → GPU device hang (regression introduced in 0.2.83)

#840 [Feature Request] Speculative decoding (draft model + batched verification) in the web runtime

#847 Feature request

#810 [Model Request] Gemma 4

#644 Error: Cannot initialize runtime because of requested maxComputeWorkgroupStorageSize exceeds limit. requested=32768, limit=16384.

#753 Issue: Browser tab crashes when loading a larger model on iOS 26 (Safari)

#836 [Bug] WebLLM Engine Initialization Fails on Qualcomm Adreno GPUs with VK_ERROR_DEVICE_LOST

#839 ASCII code

#837 Uncaught (in promise) Error: A listener indicated an asynchronous response by returning true, but the message channel closed before a response was received

#741 Firefox Webpack Size Exceeds Limit

#254 Init error, Error: This model requires WebGPU extension shader-f16

#778 [Model Request] Qwen3.5 support — Gated DeltaNet + MoE architecture

#707 Roadmap

#786 Model request: Granite 4.0

#831 Don't mind this, I realized this is a duplicate

#829 [Bug] OpenAI response created timestamps use milliseconds

#719 Support for embedding gemma

#828 what 1984 Right Think "suggestions" are inflicted?

#826 Docs: advanced usage cache examples use invalid appConfig shape

#711 docs: WebLLM model provider for the Vercel AI SDK

#824 Request models from the Gemma 4 Family