← Benchmark report cards
Agentic
Terminal-Bench
Published by Terminal-Bench (Laude Institute). Source page ↗ · Data we read (file download) ↗
Terminal-Bench 2.x as transcribed by Epoch; saturated and superseded by 4.0.
Facts
| Fact | Value | Where it comes from |
|---|---|---|
| Who runs the evaluation | an independent organisation | curated by aiqbench, reviewed 2026-10-09 |
| Test set exposure | public | curated by aiqbench, reviewed 2026-10-09 |
| Data licence | CC-BY-4.0 | the board's attribution record |
| Newest published result | May 14, 2026 | the ingested rows |
| Last pulled by aiqbench | 19 minutes ago · every 12 h | the sync log |
| Models on this axis | 202 | the ingested rows |
| Rows with a published interval | 95.5% of rows | the ingested rows |
| Headroom at the top | Best 84.7%, 15.3% below the ceiling; top 10 within 6.3%. | computed from the ingested rows |
| Axes it tracks closely |
| Spearman ρ over shared models (see Clusters) |
Last frontier jumps
- Sep 13, 2026 · gpt-5.5 · 84.7%
- Sep 13, 2026 · claude-opus-4-6 · 79.8%
- Dec 11, 2025 · gpt-5.2-2025-12-11 · 64.9%
Where it appears
Weighted in the built-in templates: Coding-focused ×2, Agentic ×3.
View this axis alone in the console →Attribution
Source: Epoch AI, 'AI Benchmarking Hub'. Published online at epoch.ai. Retrieved from https://epoch.ai/benchmarks. Licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Modified by aiqbench: normalised into aiqbench's common board format; we ingest the 'terminalbench_external.csv' table and rescale its 0-1 scores to percent. Underlying results published by Terminal-Bench (Laude Institute) — https://www.tbench.ai/leaderboard, transcribed by Epoch AI.