- Parameters
- 26.9B
- Quant
- PQ2_0 (2.13 bpw, group-128)
- Size
- 7.2 GB
- Family
- bonsai
ternary-bonsai-27b-pq2
Measured, not seateddenseresidentllama.cpp (PrismML fork)
Model card ↗Outcome as of 20 September 2026, from the seat registry. The write-up below was made when the model was tested.
Why it was tested
prism-ml's group-128 ternary pack, built for their own llama.cpp fork's hybrid-attention kernels. The upstream Q2_g64 variant lacks them.
What came of it
The only model on this board at 100% toolcall. Also the best tagger of the small seats (96.2/97.9 at n=240) and best on contract (75.0). Weak where it matters for a driver: nullschema 27.8, worldview 42.5, and 0/8 on superseded. A tool executor, not a thinker. 17.3 tok/s -- dense, so it reads all 26.9B parameters per token.
Derived from this: Bonsai-27B (Q1_0).
No runs recorded against this identifier.
- Bonsai-27B (Q1_0)Rejected
26.9B · dense · Q1_0 (1.13 bpw) · 3.5 GB
Same base as the ternary build at a more aggressive operating point. Tested to find which capabilities survive 1-bit.
- Bonsai-8B (Q1_0)Rejected
8.0B · dense · Q1_0 (1.13 bpw) · 1.1 GB
Derived from Qwen3-8B, not the Qwen3.6-27B backbone the larger Bonsai models use. Tested as the small corner of a size-by-bit-width grid.
- Ternary-Bonsai-27B (prism-ml)Measured, not seated
27B · dense · Q2_g64 (1.71 bits/weight) · 7.1 GB
End-to-end ternary weights on a Qwen3.6-27B backbone with ~75% linear attention, claiming 262K context and 94.6% of FP16 intelligence retained. 11.7 GB resident at a 128k window - a third of K2-36B for the same context.