[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"tool-pulsar":3},{"tool":4,"categoryPool":38},{"id":5,"name":6,"slug":6,"description":7,"url":8,"githubUrl":8,"logoUrl":-1,"openSource":9,"categories":10,"repo":15,"company":-1,"articles":36,"createdAt":37},"entity_01kxz8e5fqe4eb13h7ybkbm0ym","pulsar","SSD-streaming inference engine for giant MoE models (Rust + CUDA). GLM 5.2 743B at 2 tok\u002Fs and Hy3 295B at 7 tok\u002Fs on two consumer 16GB GPUs. Zero-config multi-GPU: measures PCIe bandwidth, places attention and hot experts where they fit.","https:\u002F\u002Fgithub.com\u002Fgiannisanni\u002Fpulsar",true,[11],{"id":-1,"name":12,"slug":13,"count":14},"Local Inference","local-inference",0,{"fullName":16,"url":8,"stars":17,"forks":18,"openIssues":19,"language":20,"license":-1,"topics":21,"pushedAt":34,"archived":35},"giannisanni\u002Fpulsar",205,25,4,"Rust",[22,23,24,25,26,27,28,29,30,31,32,33],"cuda","gguf","glm","inference-engine","llm","local-llm","mixture-of-experts","moe","multi-gpu","quantization","rust","ssd-streaming","2026-08-11T16:10:04Z",false,[],1784534210,[39,63,81,99],{"id":40,"name":41,"slug":42,"description":43,"url":44,"githubUrl":45,"logoUrl":-1,"openSource":9,"categories":46,"repo":48,"company":-1,"articles":61,"createdAt":62},"entity_01kwn4npgyf25rdkpzwh66v1kr","Aithy","aithy","Aithy is a private local AI runtime for state-of-the-art agent research, local inference, sandboxed tools, durable memory, and LAN Mesh resource sharing.","https:\u002F\u002Faithy.dev\u002F","https:\u002F\u002Fgithub.com\u002Fdosco\u002Faithy",[47],{"id":-1,"name":12,"slug":13,"count":14},{"fullName":49,"url":45,"stars":50,"forks":51,"openIssues":52,"language":53,"license":54,"topics":55,"pushedAt":60,"archived":35},"dosco\u002Faithy",107,7,1,"TypeScript","Apache-2.0",[56,57,26,58,59],"ai","ai-agent","local-ai","personal-ai","2026-07-16T08:25:13Z",[],1783120976,{"id":64,"name":65,"slug":66,"description":67,"url":68,"githubUrl":68,"logoUrl":-1,"openSource":9,"categories":69,"repo":71,"company":-1,"articles":79,"createdAt":80},"entity_01kxz8h4cxe4eb13j1t9zbqnfc","Colibri","colibri","Run GLM-5.2 (744B MoE) on a 25GB-RAM consumer machine — pure C, zero deps, experts streamed from disk. Tiny engine, immense model. 🐦","https:\u002F\u002Fgithub.com\u002FJustVugg\u002Fcolibri",[70],{"id":-1,"name":12,"slug":13,"count":14},{"fullName":72,"url":68,"stars":73,"forks":74,"openIssues":75,"language":76,"license":54,"topics":77,"pushedAt":78,"archived":35},"JustVugg\u002Fcolibri",24369,2658,125,"C",[],"2026-08-12T15:34:41Z",[],1784534307,{"id":82,"name":83,"slug":84,"description":85,"url":86,"githubUrl":86,"logoUrl":-1,"openSource":9,"categories":87,"repo":89,"company":-1,"articles":97,"createdAt":98},"entity_01kxz8q9j7e4eb13m7d5zqk7m7","DwarfStar","ds4","DeepSeek 4 Flash and PRO local inference engine for Metal, CUDA and ROCm","https:\u002F\u002Fgithub.com\u002Fantirez\u002Fds4",[88],{"id":-1,"name":12,"slug":13,"count":14},{"fullName":90,"url":86,"stars":91,"forks":92,"openIssues":93,"language":76,"license":94,"topics":95,"pushedAt":96,"archived":35},"antirez\u002Fds4",21282,1937,469,"MIT",[],"2026-08-09T18:31:13Z",[],1784534509,{"id":100,"name":101,"slug":102,"description":103,"url":104,"githubUrl":105,"logoUrl":106,"openSource":9,"categories":107,"repo":109,"company":-1,"articles":115,"createdAt":116},"entity_01kxwfa7trfk0tcg0ej31575e7","FastMLX","fastmlx","FastMLX is a high performance production ready API to host MLX models.","https:\u002F\u002Fblaizzy.github.io\u002Ffastmlx\u002F","https:\u002F\u002Fgithub.com\u002FBlaizzy\u002Ffastmlx","https:\u002F\u002Fr2.minima.ltd\u002Forg_01jakhtww5fk6b0jmxj0qfp9rf\u002Ffavicon-5325c06d.png",[108],{"id":-1,"name":12,"slug":13,"count":14},{"fullName":110,"url":105,"stars":111,"forks":19,"openIssues":14,"language":112,"license":-1,"topics":113,"pushedAt":114,"archived":35},"Blaizzy\u002Ffastmlx",24,"Python",[],"2024-11-18T22:51:05Z",[],1784440758]