PrismML — Announcing Bonsai 27B: The First 27B-Class Model to Run on a Phone

About<br>Blog<br>News<br>Careers<br>Docs

LAUNCH

1001011100 11010 001

Back to all posts

Announcing Bonsai 27B: The First 27B-Class Model to Run on a Phone

July 14, 2026<br>PrismML

Today, we're announcing Bonsai 27B, based on Qwen3.6 27B, the new multimodal flagship of the Bonsai family and the first model of its capability class to run on a phone.<br>Our earlier releases proved that models with 1-bit and ternary weights could produce commercially useful language models. Bonsai 27B extends that frontier to a new capability tier: multi-step reasoning, structured tool calls, vision tasks, and computer-use agentic loops that stay coherent across many steps. Until today, deploying that tier locally has been impractical for a concrete reason: a 27B model occupies roughly 54GB in 16-bit precision, and even a good 4-bit build, at 18GB, is too large for a phone and for most laptops.<br>Bonsai 27B changes that. It comes in two variants:<br>Ternary Bonsai 27B uses ternary {−1, 0, +1} weights with FP16 group-wise scaling, giving a true 1.71 effective bits per weight. At 5.9 GB, it is the quality-oriented variant: it runs on an everyday laptop with the full reasoning, tool-calling, and agentic capability.<br>1-bit Bonsai 27B uses binary {−1, +1} weights with the same group-wise scaling, giving 1.125 effective bits per weight. At 3.9 GB, it is the footprint-oriented variant, which fits within the memory budget of an iPhone 17 Pro, bringing a 27B-class model onto a phone for the first time.<br>As with every Bonsai release, the low-bit representation runs end to end across the language network, embeddings, attention, MLPs, and the LM head, with no higher-precision escape hatches. Both variants are multimodal, with the vision tower shipping in a compact 4-bit form so on-device workflows can see screenshots, documents, and camera input, not just text. Bonsai 27B carries a full 262K-token context, and supports speculative-decoding, compounding the speed with lossless draft-and-verify acceleration. Everything is available today under the Apache 2.0 License.<br>Retaining the intelligence<br>Across a 15-benchmark suite spanning knowledge, reasoning, math, coding, instruction following, tool calling, and vision  (evaluated in thinking mode, where the model's full reasoning is exercised) Ternary Bonsai 27B retains 95% of the full-precision baseline, and 1-bit Bonsai 27B retains 90% .

Category (benchmarks)<br>Qwen 3.6<br>27B<br>Ternary Bonsai<br>27B<br>1-bit Bonsai<br>27B

Math (GSM8K, MATH-500, AIME25, AIME26)<br>95.3<br>93.4<br>91.7

Coding (HumanEval+, MBPP+, LiveCodeBench)<br>88.7<br>86.0<br>81.9

Agentic and Tool-calling (BFCL v3, TauBench)<br>80.0<br>74.0<br>66.0

Instruction following (IFEval, IFBench)<br>78.4<br>71.8<br>65.8

Knowledge / STEM (MMLU-Redux, MuSR)<br>83.1<br>77.0<br>73.4

Vision (MMMU Pro, OCRBench)<br>72.6<br>65.2<br>59.6

Overall (15 benchmarks)<br>85.0<br>80.5<br>76.1

‍Fig I: Benchmark scores of Bonsai 27B (thinking mode) against the full-precision baseline. Full per-benchmark results are in the whitepaper.‍<br>Read the table by capability and the story is sharper than the averages: math and coding are nearly untouched, tool calling stays within a few points of full precision - exactly the capabilities that agentic workloads depend on. For comparison, the most aggressive conventional low-bit build of the same base model scores significantly lower than 1-bit Bonsai 27B while occupying 2.5x more memory.<br>This is the same Pareto shift we demonstrated with our earlier language and image models, now at 27B scale: 27B-class capability at a footprint smaller than a full-precision 2B model. By intelligence density — the measure we introduced with 1-bit Bonsai 8B — 1-bit Bonsai 27B delivers 0.53 per GB: more than 10x the full-precision baseline, and roughly 2.7x the best low-bit alternative available.

Fig II: Intelligence density (per GB) of Bonsai 27B compared to other models in the same parameter class.Why this is an important paradigm shift<br>The most valuable AI workloads are shifting from single responses to sustained work: assistants that operate real tools, workflows that run unattended before returning a result, and research that synthesizes dozens of documents. This shift changes the shape of the workload — an agent doesn't make one model call, it makes hundreds, each one carrying context, producing structured output, and feeding the next.<br>Cloud APIs will remain the right choice for many products. But for agentic workloads, cloud-only execution imposes structural constraints: every step is a remote request, per-token cost accumulates with every iteration, and every plan, tool call, and intermediate result crosses the network including the user's private files, screen, and data.

Carousel I: End-to-end agentic workflow with Hermes, powered by our Ternary Bonsai 27B model on NVIDIA GeForce RTX 5090.

Carousel II: Agentic tool calling and MCP integration with Ternary Bonsai 27B on M5 Max.

Local execution changes the equation. When a model...