Ternary Bonsai 2 27B

Name: Ternary Bonsai 2 27B
Author: PrismML

Reasoning model

Also known as: Bonsai 2 27B

Release date: 2026-09-17Views: 0

Live demoGitHub Hugging Face Compare

Parameters

27.36B

Context length

256K

Multilingual

No data

Reasoning ability

No data

Ternary Bonsai 2 27B is an AI model published by PrismML, released on 2026-09-17, for Reasoning model, with 27.36B parameters, and 256K context length, requiring about 5.95 GB (GGUF PTQ1_0, language weights only) storage, with a 98.80 score on MATH-500.

Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology

Ternary Bonsai 2 27B

Model basics

Reasoning traces

Supported

Thinking modes

Thinking Level · Extra-High (Default)Standard ModeThinking Level · Medium

Context length

256K tokens

Max output length

No data

Model type

Reasoning model

Modality (in / out)

Text, Image → Text

Release date

2026-09-17

Model file size

5.95 GB (GGUF PTQ1_0, language weights only)

MoE architecture

Total params / Active params

27.36B / Not applicable

Knowledge cutoff

No data

Ternary Bonsai 2 27B

Open source & experience

Code license

Apache 2.0

Weights license

Apache 2.0- Commercial use permitted

GitHub repo

github.com/PrismML-Eng/Bonsai-demo

Hugging Face

huggingface.co/prism-ml/Ternary-Bonsai-2-27B-gg…

Live demo

N/A

Ternary Bonsai 2 27B

Official resources

Paper

DataLearnerAI blog

N/A

Ternary Bonsai 2 27B

API details

API speed

No data

No public API pricing yet.

Ternary Bonsai 2 27B

Benchmark Results

Ternary Bonsai 2 27B currently shows benchmark results led by GSM8K (1 / 71, score 96.66), MATH-500 (1 / 46, score 98.80), LiveCodeBench (10 / 251, score 90.07). This page also consolidates core specs, context limits, and API pricing so you can evaluate the model from benchmark results and deployment constraints together.