BOBonsai 27B
Bonsai 27B: Ultra-Efficient Sparse Frontier LLM
Bonsai 27B is PrismML's 27-billion parameter sparse mixture-of-experts model engineered to deliver frontier-class reasoning with 4× lower memory and inference footprint.
🌲 Architecture & MoE Efficiency
- Total Parameters: 27B parameters.
- Active Parameters: 4.1B active parameters per token.
- Context Window: 128k tokens with flash-attention support.
📊 Benchmark Results
| Benchmark | Task Domain | Bonsai 27B | Dense 70B Baseline |
|---|---|---|---|
| MMLU | General Knowledge | 82.4% | 81.9% |
| GSM8K | Math Reasoning | 88.6% | 86.2% |
| HumanEval | Code Generation | 84.1% | 81.5% |
Key Features
Ternary (1.71 effective bits) and 1-bit (1.125 effective bits) variants with no higher-precision escape hatches
Multimodal with a compact 4-bit vision tower
Supports multi-step reasoning, structured tool calls, and computer-use agentic loops
Native support on Apple devices via MLX and NVIDIA GPUs via CUDA
Supports speculative decoding
You might also want to compare
Verified Sources
Tags
Model Specs
Parameters
27B
Context Window
262K tokens
License
Apache 2.0
Deployment
Resources & Links
Lineage
Model Family
Part of the bonsai family
Only release in this line currently tracked.
Compare Specs
Compare parameters, context windows, modalities, and benchmark scores of this model side-by-side with others.
Compare Model