A display-only Artificial Analysis coding metric for agentic coding and terminal use on a harder Terminal-Bench slice.
BenchLM mirrors the published score view for Terminal-Bench Hard. Grok 4.3 leads the public snapshot at 37.9%. BenchLM does not use these results to rank models overall.
Year
2026
Tasks
Agentic coding and terminal tasks
Format
Task success rate
Difficulty
Professional software engineering
BenchLM stores Terminal-Bench Hard separately from Terminal-Bench 2.0 because OpenRouter and Artificial Analysis publish it as a distinct benchmark card.
Version
Terminal-Bench Hard 2026
Refresh cadence
Quarterly
Staleness state
Current
Question availability
Public benchmark set
BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.
A display-only Artificial Analysis coding metric for agentic coding and terminal use on a harder Terminal-Bench slice.
Grok 4.3 by xAI currently leads with a score of 37.9% on Terminal-Bench Hard.
1 AI models have been evaluated on Terminal-Bench Hard on BenchLM.
For engineers, researchers, and the plain curious — a weekly brief on new models, ranking shifts, and pricing changes.
Free. No spam. Unsubscribe anytime.