Model results

LiveCodeBench

Choose a shared evaluation protocol to compare model scores. Versions, sources and configurations stay separate.

Atlas snapshot captured 10/19/2026, 10:20:28 AM UTC. This is an import timestamp, not the evaluation date. Coverage may be incomplete.

Reviewed recommendation and limits

Choose dated LiveCodeBench windows for contest algorithms. Use Pro V2 for repository issues and Terminal-Bench for terminal work.

Dated code-generation, repair, execution and output-prediction tasks. Imported groups below may use different protocols; this review does not certify each run. Match the problem window to the model cutoff and retain the dataset revision. Official errata documents grading and test defects. An unknown training cutoff prevents a strong contamination-free claim.

Publisher results LiveCodeBench research LiveCodeBench grading errata

← All evaluation groups

LiveCodeBench

Source: PrismML model card (vendor-reported)
Version: unversioned
Measurement date: 2026-09-17
Metric: % · Higher is better
Evaluation scope: Vendor-reported; thinking mode; EvalScope + vLLM on NVIDIA H100; temperature 1.0, top_p 0.95, top_k 20; result date not stated by the publisher, dated to the model-card publication; same-run FP16 Qwen3.8-27B reference: 90.05

Positions are withheld because at least one result lacks explicit independent provenance or a documented evaluation scope. Missing metadata is shown as unreported; matching imported metadata is not independent verification of the protocol.

Ternary Bonsai 2 27B
prism-ml/ternary-bonsai-2-27b-20260918
90.07%
View modelEvidence model-card-reported