{"version":1,"type":"story","url":"https://digestai.news/story/agentionai-releases-qwen3-8-flash-next-gyro-coder-for-24-gb-gpus","json":"https://digestai.news/story/agentionai-releases-qwen3-8-flash-next-gyro-coder-for-24-gb-gpus.json","markdown":"https://digestai.news/story/agentionai-releases-qwen3-8-flash-next-gyro-coder-for-24-gb-gpus.md","slug":"agentionai-releases-qwen3-8-flash-next-gyro-coder-for-24-gb-gpus","headline":"AgentionAI releases Qwen3.8-Flash-Next-Gyro-Coder for 24 GB GPUs","summary":"AgentionAI has published Qwen3.8-Flash-Next-Gyro-Coder, a quantized version of the Qwen3.8-Flash-Next model designed to run on a single consumer GPU with 24 GB of memory, such as an RTX 3090, 4090, or 7900 XTX. The model utilizes a mixture-of-experts architecture, retaining only the 256 experts out of 512 that are relevant for programming tasks. This selective pruning, combined with a custom trellis encoder, reduces the GPU memory footprint to 19.9 GiB while maintaining high fidelity for coding tasks.\n\nThe model achieves 93% top-1 agreement with the full model on held-out Python data and passed 13 of 16 coding tests in the author's evaluation. A larger variant, Gyro-Coder-L, is available for 32 GB cards, using 27.5 GiB of memory and passing 15 of 16 tests. The release includes instructions for running the model via a custom llama.cpp Docker build, supporting both NVIDIA and AMD/Intel GPUs. The file size is 70.6 GiB on disk due to an n-gram table stored on NVMe storage.\n\nAgentionAI notes that the model is best suited for clear, well-specified coding instructions like implementing or refactoring code, rather than open-ended general knowledge queries. The project is currently experimental, with potential changes to file formats and kernels expected in future updates.","keyPoints":["Runs on 24 GB GPUs like RTX 3090 using 19.9 GiB of memory.","Retains 256 of 512 experts selected via evolutionary search for coding.","Passes 13 of 16 coding tests; 93% top-1 agreement on Python."],"whyItMatters":"Enables high-performance coding assistance on affordable consumer hardware, lowering the barrier to entry for local LLM development and reducing reliance on cloud APIs for code generation tasks.","category":{"slug":"models","name":"Generative AI & Models","url":"https://digestai.news/category/models"},"entities":{"companies":["AgentionAI"],"models":["Qwen3.8-Flash-Next","Gyro-Coder","Gyro-Coder-L","Gyro-M","Gyro-S"],"people":[]},"firstPublishedAt":"2026-10-23T16:16:38Z","updatedAt":"2026-10-23T16:16:38Z","sourceCount":1,"hasPrimarySource":true,"sources":[{"outlet":"huggingface.co","title":"agentionai/Qwen3.8-Flash-Next-Gyro-Coder-GGUF · Hugging Face","url":"https://huggingface.co/agentionai/Qwen3.8-Flash-Next-Gyro-Coder-GGUF","publishedAt":"2026-10-23T16:16:38Z","type":"primary","primary":true,"lead":true}],"sourceNotes":null,"discussions":[{"site":"Reddit","url":"https://www.reddit.com/r/LocalLLaMA/comments/1xdihql/agentionaiqwen38flashnextgyrocodergguf_hugging/","points":null}],"thread":{"title":"Qwen Flash Next Hardware Optimization Race","url":"https://digestai.news/thread/inco-ai-releases-dflash-2-draft-model-for-qwen3-8-flash-next","storyCount":3},"cite":{"text":"Digest AI, \"AgentionAI releases Qwen3.8-Flash-Next-Gyro-Coder for 24 GB GPUs\", 23 October 2026, https://digestai.news/story/agentionai-releases-qwen3-8-flash-next-gyro-coder-for-24-gb-gpus","publisher":"Digest AI","title":"AgentionAI releases Qwen3.8-Flash-Next-Gyro-Coder for 24 GB GPUs","datePublished":"2026-10-23T16:16:38Z","url":"https://digestai.news/story/agentionai-releases-qwen3-8-flash-next-gyro-coder-for-24-gb-gpus"},"generatedBy":"Written by Digest AI's editorial model from the linked sources; the sources are the record.","license":"Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse"}