cd /news/large-language-models/tested-bonsai-2-27b-locally-ternary-… · home topics large-language-models article
[ARTICLE · art-133836] src=dotnetperls.com ↗ pub= topic=large-language-models verified=true sentiment=↑ positive

Tested Bonsai 2 27B Locally (Ternary Model)

A tester ran Bonsai 2 27B, a ternary model that stores each weight as one of three values (1, 0, -1), locally on a 12GB Nvidia GPU using a special build of llama-cpp from PrismML's GitHub. The 7.2 GB model, based on Qwen 3.8 27B, decoded at roughly 30 tokens per second on an RTX 3060 with 12 GB of RAM and worked as an agentic coder that rarely fails tool calls, though it tends to over-think like the original Qwen 3.8. The tester reported Bonsai 2 27B is a significant quality improvement over previous Bonsai 27B models and that the claim it is only 2% degraded from Qwen 3.8 27B may be at least somewhat accurate in general.

read1 min views1 publishedSep 18, 2026

Ternary models represent each weight as one of 3 values (1, 0, and -1). They can be much smaller and faster than other quantizations but this comes at a price of quality. I downloaded the Bonsai 2 model (which is 7.2 GB) and used it in Pi code.

First, a special version of llama-cpp is currently needed and can be downloaded from PrismML's GitHub. I have a 12GB Nvidia GPU so the 7.2 GB model fits nicely; the 4-bit quantization of Qwen 3.8 would not fit. Bonsai 2 is based on Qwen 3.8 27B, a popular local coding model.

I found:

• Bonsai 2 27B works as an agentic coder, and it rarely fails tool calls, although it tends to over-think like the original Qwen 3.8 model does.

• Bonsai 2 27B is a significant improvement over the previous Bonsai 27B models in quality of output.

• The claim that Bonsai 2 27B is only 2% degraded from Qwen 3.8 27B may be at least somewhat accurate in general.

• On an RTX 3060 desktop card with 12 GB RAM the model decodes at around 30 tokens per second, which is usable (although the fan is loud).

• Bonsai 2 is useful for real coding tasks and may help alleviate the need to buy GPUs with huge amounts of memory.

── more in #large-language-models 4 stories · sorted by recency
── more on @bonsai 2 27b 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
Live at https://your-agent.zahid.host
Get free account → Pricing
from €0/mo · no card required
LIVE [news/tested-bonsai-2-27b-…] indexed:0 read:1min 2026-09-18 ·