{"full_name":"AlexsJones/llmfit","owner":"AlexsJones","name":"llmfit","url":"https://github.com/AlexsJones/llmfit","description":"Hundreds of models & providers. One command to find what runs on your hardware.","homepage":"","topics":[],"language":"Rust","license":"MIT","avatar":"https://avatars.githubusercontent.com/u/1235925?u=b5a85e9e4d3ca4bba85a7c37bc63d9d968cf1b0c&v=4","created_at":"2026-02-15T16:11:31Z","pushed_at":"2026-11-09T21:31:22Z","metrics":{"stars":38869,"forks":2511,"watchers":114,"issues":292,"prs":924,"releases":138,"issues_open":null,"issues_closed":null,"prs_open":null,"prs_closed":null,"prs_merged":null,"discussions":23},"growth":{"day":{"stars":62,"forks":5,"base_day":"2026-11-09"}},"history":{"dates":["2026-11-09","2026-11-10"],"stars":[38807,38869],"forks":[2506,2511],"watchers":[114,114],"issues_open":[null,null],"prs_open":[null,null]},"boards":{},"same_owner":[],"issues":[{"number":638,"title":"Detect multiple GPUs","comments":12,"created_at":"2026-06-17T14:08:08Z"},{"number":301,"title":"Request to add support for NPU and Hybrid models","comments":10,"created_at":"2026-03-31T08:53:56Z"},{"number":81,"title":"Feature: collect results on a website","comments":8,"created_at":"2026-02-25T08:22:42Z"},{"number":914,"title":"[Feature]: Recommendations for Tool-Using Coding Agents","comments":7,"created_at":"2026-08-19T12:55:12Z"},{"number":635,"title":"MacBook Neo shows no models under \"Good\" or \"Runnable\" due to conservative KV cache filtering","comments":7,"created_at":"2026-06-15T23:28:30Z"},{"number":1090,"title":"[Calibration]: Detect PCIe link speed/topology for accurate calibration and throughput estimates","comments":4,"created_at":"2026-09-27T20:28:46Z"},{"number":1036,"title":"Fit estimates don't account for Unsloth Studio's beyond-VRAM offload","comments":4,"created_at":"2026-09-14T04:48:01Z"},{"number":928,"title":"[Feature]: Support LM Studio in Inference Bench (Quality Benchmarks / Routing)","comments":4,"created_at":"2026-08-21T09:21:39Z"}],"discussions":[{"number":1178,"title":"llmfit is #1 in Open models on Archestack (badge, if you want it)","comments":0,"category":"Show and tell","created_at":"2026-10-21T15:30:19Z"},{"number":1154,"title":"llmfit included in AgentManager's tool catalog","comments":1,"category":"Show and tell","created_at":"2026-10-18T10:21:17Z"},{"number":1134,"title":"Optional Packslip CLI manifests and release-matched advisor skill","comments":2,"category":"Ideas","created_at":"2026-10-11T17:11:12Z"},{"number":1025,"title":"abliterated/community models work?","comments":4,"category":"Q&A","created_at":"2026-09-11T14:52:45Z"},{"number":989,"title":"LLM Fit GUI — a Windows desktop wrapper for llmfit","comments":1,"category":"Show and tell","created_at":"2026-09-02T16:35:32Z"},{"number":956,"title":"Export/Print","comments":1,"category":"Ideas","created_at":"2026-08-26T15:03:52Z"},{"number":949,"title":"Filtering Context size , filtering Quants, filtering Tps","comments":1,"category":"Ideas","created_at":"2026-08-24T09:46:39Z"},{"number":929,"title":"Possible to filter by nationality of model?","comments":4,"category":"Q&A","created_at":"2026-08-21T15:09:01Z"}],"latest_release":{"name":"v1.1.20","tag":"v1.1.20","published_at":"2026-11-03T17:39:35Z","body":"## [1.1.20](https://github.com/AlexsJones/llmfit/compare/v1.1.19...v1.1.20) (2026-11-03)\n\n\n### Bug Fixes\n\n* NVFP4 and MXFP4 safetensors repos are no longer offered as llama.cpp fits, so they don't top the list on ROCm and CPU machines ([#1230](https://github.com/AlexsJones/llmfit/issues/1230))\n* 39 large models now have real attention-head counts from their `config.json` (including Llama-3.1-405B and MiniMax-M1), so `--cluster` can check their tensor-parallel splits ([#1230](https://github.com/AlexsJones/llmfit/issues/1230))\n* `diff` and the TUI compare view no longer put a model that doesn't fit, or whose cluster split is refused, ahead of one that fits ([#1230](https://github.com/AlexsJones/llmfit/issues/1230))\n* **cluster:** the banner shows the VRAM pool actually used when `--memory` is set ([#1230](https://github.com/AlexsJones/llmfit/issues/1230))\n* **cluster:** Ray discovery sets `backend` from the GPU vendor where Ray reports it, and otherwise names the nodes that need `backend` set instead of assuming CUDA ([#1230](https://github.com/AlexsJones/llmfit/issues/1230))\n* **cluster:** `--json` errors from `cluster init`, `cluster status`, `cluster clear` and `--cluster`; \"1 node\" instead of \"1 nodes\"; `cpu_cores` taken from the machine's core count, not Ray's `CPU` resource ([#1230](https://github.com/AlexsJones/llmfit/issues/1230))\n\n## What's Changed\n* fix: v1.1.19 Framework/ROCm findings: NVFP4 routing, head backfill, cluster diff/banner/Ray backend by @AlexsJones in https://github.com/AlexsJones/llmfit/pull/1230\n* chore(main): release 1.1.20 by @AlexsJones in https://github.com/AlexsJones/llmfit/pull/1231\n\n\n**Full Changelog**: https://github.com/AlexsJones/llmfit/compare/v1.1.19...v1.1.20"},"readme":true,"readme_zh":true,"news":[],"stale":false,"snapshot_at":"2026-11-10T09:35:25Z","history_start":"2026-09-06"}