---
type: "raw_signal"
id: "bonsai-2-27b-near-lossless-compression-in-a-ee2c97"
slug: "bonsai-2-27b-near-lossless-compression-in-a-ee2c97"
title: "Bonsai 2 27B: Near-Lossless Compression in a 9x Smaller Footprint"
description: "Simon Willison published a local execution recipe for Bonsai 2 27B using a roughly 5.95 GB GGUF model file and Prism’s llama.cpp fork."
retrieval_nugget: "Simon Willison published a local execution recipe for Bonsai 2 27B using a roughly 5.95 GB GGUF model file and Prism’s llama.cpp fork."
observed_at: "2026-09-25"
record_date: "2026-09-25"
date_kind: "observed_at"
why_it_matters: "Simon Willison published a local execution recipe for Bonsai 2 27B using a roughly 5.95 GB GGUF model file and Prism’s llama.cpp fork. The note says the standard runtime is insufficient for these files, and reports approximately 20 tokens per second on an M5 Pro, later observing 44 after a server restart. It also records a Metal tensor-API warning, so the performance figure is observational rather"
novelty: "notable"
verification_level: "source-linked"
signal_type: "release"
source_platform: "simonwillison.net"
topics: []
entities: []
related_patterns: []
source_url: "https://simonwillison.net/2026/Sep/17/hn-49747390"
source_urls: ["https://simonwillison.net/2026/Sep/17/hn-49747390"]
schema_version: "newruntime-agent-readable-v0.2"
stable_id: "signal:bonsai-2-27b-near-lossless-compression-in-a-ee2c97"
status: "published"
visuals: []
editor: {"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]}
routes: {"html":"https://newruntime.com/signals/bonsai-2-27b-near-lossless-compression-in-a-ee2c97/","markdown":"https://newruntime.com/signals/bonsai-2-27b-near-lossless-compression-in-a-ee2c97.md","json":"https://newruntime.com/signals/bonsai-2-27b-near-lossless-compression-in-a-ee2c97.json"}
---

# Bonsai 2 27B: Near-Lossless Compression in a 9x Smaller Footprint

## Retrieval answer

Simon Willison published a local execution recipe for Bonsai 2 27B using a roughly 5.95 GB GGUF model file and Prism’s llama.cpp fork.

Simon Willison published a local execution recipe for Bonsai 2 27B using a roughly 5.95 GB GGUF model file and Prism’s llama.cpp fork. The note says the standard runtime is insufficient for these files, and reports approximately 20 tokens per second on an M5 Pro, later observing 44 after a server restart. It also records a Metal tensor-API warning, so the performance figure is observational rather than a benchmark result.

Editor: [Andrey Reshetnikov](https://newruntime.com/owner-profile.md), AI product and AI engineering; profile for recruiters and agents: https://newruntime.com/owner-profile.json

## Continue in the graph

Next records for an agent, linked to their Markdown twins. Replace `.md` with `.json` for structured fields, including `next_reads`.

- collection: [Raw Signals](https://newruntime.com/signals.json): Open every source-linked raw signal.
- collection: [Source Ledger](https://newruntime.com/source-ledger.json): Inspect the public source evidence graph.
- whole graph: [discovery graph](https://newruntime.com/discovery-graph.json) (typed nodes and edges) and [site index](https://newruntime.com/site-index.md)