GitHub discovery

Promising new AI projects

Young, active repositories worth an early look.

How this list is chosen

Candidates are the repositories GitHub's topic search returns for "llm" and "local-llm" when we collect, about three times a day. A repository must have been observed in the last 3 days to be listed, and each view lists at most 12.

Figures as of our latest count, . We have stored star counts for 100 repositories since .

  1. bisand/kvad

    Local AI for Apple silicon, written from the arithmetic up: an LLM, image and video inference engine in Rust with Metal kernels for the M5, a web UI, a terminal app and an OpenAI-compatible API.

    Licence
    MIT
    Last activity
    Last push
    Latest release
    v0.20.0,
    Stars
    0 on
    Growth
    Not measured yet: we have counted its stars once, and have been measuring it since .

    Why look

    • Shipped release v0.20.0 on 17 Oct 2026.
    • Created 19 Sept 2026, last push 17 Oct 2026.
    • Tagged apple-silicon, flux, inference, llm.

  2. Pigeon-Home/homing-pigeon

    Self-hosted, SI-controlled local agent. Tell your cloud SI (Grok, ChatGPT, etc.) to send signed jobs via email or a cloud queue, and they execute on your machine: Linux, macOS, FreeBSD, Windows. An Oracle on the host makes sure nothing bad or unintended happens.

    Licence
    Apache-2.0
    Last activity
    Last push
    Latest release
    v3.7.22,
    Stars
    1 on
    Growth
    Not measured yet: we have counted its stars once, and have been measuring it since .

    Why look

    • Shipped release v3.7.22 on 17 Oct 2026.
    • Created 13 Oct 2026, last push 17 Oct 2026.
    • Tagged ai-agents, automation, email, intrusion-detection.

  3. trafficflowhq/ohmail

    Consent-first email on your existing IMAP mailboxes. Optional AI assistet screener with local or Cloud models. Full local Clients for MacOS, Windows, Linux + Arch and Omarchy integrationst. Android + iOS (coming soon) Apps. AGPL-3.0.

    Licence
    AGPL-3.0
    Last activity
    Last push
    Latest release
    v0.27.0,
    Stars
    83 on
    Growth
    Not measured yet: we have counted its stars once, and have been measuring it since .

    Why look

    • Shipped release v0.27.0 on 17 Oct 2026.
    • Created 30 Jul 2026, last push 17 Oct 2026.
    • Tagged ai, consent-first, email, imap.

  4. teng-lin/open-foresight

    Offline AI meeting notes and transcription for Linux and macOS. Open-source Google AI Edge Foresight and Granola alternative with local Gemma LLM, Whisper, RAG and MCP.

    Licence
    MIT
    Last activity
    Last push
    Latest release
    v0.1.0,
    Stars
    1 on
    Growth
    Not measured yet: we have counted its stars once, and have been measuring it since .

    Why look

    • Shipped release v0.1.0 on 17 Oct 2026.
    • Created 16 Oct 2026, last push 17 Oct 2026.
    • Tagged ai-meeting-assistant, gemma, granola-alternative, knowledge-base.

  5. nfekete82/mlx-nobby

    Local AI workspace for Apple Silicon powered by MLX — chat, agents, RAG, image generation and more.

    Licence
    MIT
    Last activity
    Last push
    Latest release
    v1.19.0,
    Stars
    0 on
    Growth
    Not measured yet: we have counted its stars once, and have been measuring it since .

    Why look

    • Shipped release v1.19.0 on 17 Oct 2026.
    • Created 10 Sept 2026, last push 17 Oct 2026.
    • Tagged ai-assistants, apple-silicon, docker, fastapi.

  6. Thomas-Pietsch/dudes

    Local, spec-first AI software delivery with a human sign-off, recorded in git. Agents deliver. You seal.

    Licence
    Apache-2.0
    Last activity
    Last push
    Latest release
    None published (checked )
    Stars
    0 on
    Growth
    Not measured yet: we have counted its stars once, and have been measuring it since .

    Why look

    • Created 15 Oct 2026, last push 17 Oct 2026.
    • Tagged ai-agents, aider, golang, human-in-the-loop.

  7. tejas4patel/dragonfarm

    Raise your own dragon: fine-tune a language model on your own documents, from R. Writes the training data, trains on your laptop, RAG built in.

    Licence
    A licence GitHub could not identify — read it
    Last activity
    Last push
    Latest release
    v0.3.4,
    Stars
    13 on
    Growth
    Not measured yet: we have counted its stars once, and have been measuring it since .

    Why look

    • Shipped release v0.3.4 on 16 Oct 2026.
    • Created 14 Sept 2026, last push 17 Oct 2026.
    • Tagged fine-tuning, llm, local-llm, lora.

  8. gloutchov/LocalAgentBenchmark

    Benchmark locale, ripetibile e offline per confrontare modelli Ollama come coding agent tramite Pi

    Licence
    Apache-2.0
    Last activity
    Last push
    Latest release
    v0.12.0,
    Stars
    0 on
    Growth
    Not measured yet: we have counted its stars once, and have been measuring it since .

    Why look

    • Shipped release v0.12.0 on 17 Oct 2026.
    • Created 9 Sept 2026, last push 17 Oct 2026.
    • Tagged benchmark, coding-agents, local-llm, ollama.

  9. exclude-barrier/OrsikTop

    OrsikTop is a fast Rust TUI for monitoring local LLM inference.

    Licence
    Apache-2.0
    Last activity
    Last push
    Latest release
    v0.4.1,
    Stars
    0 on
    Growth
    Not measured yet: we have counted its stars once, and have been measuring it since .

    Why look

    • Shipped release v0.4.1 on 17 Oct 2026.
    • Created 10 Sept 2026, last push 17 Oct 2026.
    • Tagged amd, dashboard, gpu-monitoring, intel.

  10. Raaif-Yousuf/Wingman

    One key press, one action, one card. A lightweight Windows 11 tray assistant behind the Copilot key that looks at your screen and helps, with cloud or local (Ollama) models.

    Licence
    MIT
    Last activity
    Last push
    Latest release
    v0.2.0,
    Stars
    5 on
    Growth
    Not measured yet: we have counted its stars once, and have been measuring it since .

    Why look

    • Shipped release v0.2.0 on 17 Oct 2026.
    • Created 17 Sept 2026, last push 17 Oct 2026.
    • Tagged ai-assistant, copilot-key, desktop-app, local-llm.

  11. Marco-Lu2508/fLLM

    Local LLM inference engine in pure Rust: GGUF, OpenAI/Anthropic/Ollama-compatible API, CUDA (incl. NVFP4), Metal, Vulkan, CPU

    Licence
    Apache-2.0
    Last activity
    Last push
    Latest release
    None published (checked )
    Stars
    0 on
    Growth
    Not measured yet: we have counted its stars once, and have been measuring it since .

    Why look

    • Created 14 Oct 2026, last push 17 Oct 2026.
    • Tagged cuda, gguf, huggingface, llm.

  12. 1bit-MONSTER/engine

    1bit engine: local LLM inference for AMD Ryzen AI (Strix Halo) inside Lemonade. HRX with our Loom kernels on the Radeon iGPU, plus the XDNA 2 NPU; OpenAI-compatible API. Runs Qwen3.8, Zyphra ZAYA1, ternary Bonsai.

    Licence
    Apache-2.0
    Last activity
    Last push
    Latest release
    None published (checked )
    Stars
    5 on
    Growth
    Not measured yet: we have counted its stars once, and have been measuring it since .

    Why look

    • Created 22 Sept 2026, last push 17 Oct 2026.
    • Tagged amd, gfx1151, gguf, gpt-oss.