> ## Documentation Index
> Fetch the complete documentation index at: https://docs.prismml.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Community & Resources

> Whitepapers, source repositories, community benchmarks, and where to get help.

## Technical references

<CardGroup cols={2}>
  <Card title="Bonsai collection" icon="cube" href="https://huggingface.co/collections/prism-ml">
    All Bonsai models repositories: 1-bit and ternary, GGUF and MLX.
  </Card>

  <Card title="1-bit and Ternary Bonsai 27B whitepaper" icon="file-pdf" href="https://github.com/PrismML-Eng/Bonsai-demo/blob/main/1-bit-bonsai-27b-whitepaper.pdf">
    Architecture, training approach, and evaluation methodology for the 27B family.
  </Card>

  <Card title="1-bit Bonsai 8B whitepaper" icon="file-pdf" href="https://github.com/PrismML-Eng/Bonsai-demo/blob/main/1-bit-bonsai-8b-whitepaper.pdf">
    Architecture, training approach, and evaluation methodology for the 1-bit 8B.
  </Card>

  <Card title="Ternary-Bonsai 8B whitepaper" icon="file-pdf" href="https://github.com/PrismML-Eng/Bonsai-demo/blob/main/ternary-bonsai-8b-whitepaper.pdf">
    The 1.58-bit representation, group-wise scaling, and benchmark results.
  </Card>
</CardGroup>

## Source code

| Repository                                                                                       | What it is                                                                                                                                                      |
| ------------------------------------------------------------------------------------------------ | --------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| [PrismML-Eng/Bonsai-demo](https://github.com/PrismML-Eng/Bonsai-demo)                            | Setup scripts, run/server wrappers, and everything the [Quickstart](/get-started/quickstart) uses                                                               |
| [`TOOLS.md`](https://github.com/PrismML-Eng/Bonsai-demo/blob/main/TOOLS.md) (in the demo repo)   | Tool calling and MCP server setup for Bonsai 27B                                                                                                                |
| [`AGENTS.md`](https://github.com/PrismML-Eng/Bonsai-demo/blob/main/AGENTS.md) (in the demo repo) | Hardware-tuning knobs, written for AI coding agents helping someone set up the demo                                                                             |
| [PrismML-Eng/llama.cpp](https://github.com/PrismML-Eng/llama.cpp)                                | llama.cpp fork (`prism` branch) with ternary `Q2_0` kernels and [pre-built binaries](https://github.com/PrismML-Eng/llama.cpp/releases/tag/prism-b9570-0ad1dab) |
| [PrismML-Eng/mlx](https://github.com/PrismML-Eng/mlx)                                            | MLX fork with 1-bit support, until [mlx#3161](https://github.com/ml-explore/mlx/pull/3161) merges upstream                                                      |
| [huggingface.co/prism-ml](https://huggingface.co/prism-ml)                                       | All model weights                                                                                                                                               |

## Community benchmarks

Measured throughput on real hardware (RTX 3080, GB10, Strix Halo, M4 Pro, and more) is collected in [`community-benchmarks/`](https://github.com/PrismML-Eng/Bonsai-demo/tree/main/community-benchmarks) in the demo repo. To contribute numbers for your hardware, copy the llama.cpp or MLX template in that folder, run the documented procedure, and open a pull request.

## Get help

* **Discord:** [discord.gg/prismml](https://discord.gg/prismml) for questions, hardware reports, and announcements.
* **GitHub issues:** [Bonsai-demo issues](https://github.com/PrismML-Eng/Bonsai-demo/issues) for bugs in the scripts or binaries.
* **X / LinkedIn:** [@PrismML](https://x.com/PrismML) and [PrismML on LinkedIn](https://www.linkedin.com/company/prismml) for release news.