Terminal-Bench 2.0
Terminal-Bench 2.0 is an updated benchmark for testing AI agents' tool use ability to operate a computer via terminal. It evaluates how well models can handle real-world, end-to-end tasks autonomously, including compiling code, training models, setting up servers, system administration, security tasks, data science workflows, and cybersecurity vulnerabilities.
GPT-5.5 from OpenAI currently leads the Terminal-Bench 2.0 leaderboard with a score of 0.827 across 39 evaluated AI models.
GPT-5.5 leads with 82.7%, followed by 
Claude Mythos Preview at 82.0% and GPT-5.3 Codex at 77.3%.
Progress Over Time
Interactive timeline showing model performance evolution on Terminal-Bench 2.0
Terminal-Bench 2.0 Leaderboard
| Context | Cost | License | ||||
|---|---|---|---|---|---|---|
| 1 | OpenAI | — | 1.1M | $5.00 / $30.00 | ||
| 2 | Anthropic | — | — | — | ||
| 3 | OpenAI | — | 400K | $1.75 / $14.00 | ||
| 4 | OpenAI | — | 1.0M | $2.50 / $15.00 | ||
| 5 | Anthropic | — | 1.0M | $5.00 / $25.00 | ||
| 6 |  Zhipu AI | 754B | 200K | $1.40 / $4.40 | ||
| 7 |  Google | — | 1.0M | $2.50 / $15.00 | ||
| 8 |  DeepSeek | 1.6T | 1.0M | $1.74 / $3.48 | ||
| 9 |  Moonshot AI | 1.0T | 262K | $0.95 / $4.00 | ||
| 10 | Anthropic | — | 1.0M | $5.00 / $25.00 | ||
| 11 | OpenAI | — | 400K | $1.75 / $14.00 | ||
| 12 |  Alibaba Cloud / Qwen Team | — | 1.0M | $0.50 / $3.00 | ||
| 13 | OpenAI | — | 400K | $0.75 / $4.50 | ||
| 14 | Anthropic | — | 200K | $5.00 / $25.00 | ||
| 14 | Alibaba Cloud / Qwen Team | 28B | 262K | $0.60 / $3.60 | ||
| 16 | Anthropic | — | 200K | $3.00 / $15.00 | ||
| 17 |  Meta | — | — | — | ||
| 18 |  Xiaomi | 1.0T | — | — | ||
| 19 |  MiniMax | — | 205K | $0.30 / $1.20 | ||
| 20 | DeepSeek | 284B | 1.0M | $0.14 / $0.28 | ||
| 21 | Zhipu AI | 744B | 200K | $1.00 / $3.20 | ||
| 22 | Google | — | — | — | ||
| 23 | OpenAI | — | 400K | $1.25 / $10.00 | ||
| 24 | Alibaba Cloud / Qwen Team | 397B | 262K | $0.60 / $3.60 | ||
| 25 | Alibaba Cloud / Qwen Team | 35B | — | — | ||
| 26 |  StepFun | 196B | 66K | $0.10 / $0.40 | ||
| 27 | Moonshot AI | 1.0T | — | — | ||
| 28 | Alibaba Cloud / Qwen Team | 122B | 262K | $0.40 / $3.20 | ||
| 29 | Google | — | 1.0M | $0.50 / $3.00 | ||
| 30 | DeepSeek | 685B | — | — | ||
| 30 | DeepSeek | 685B | 164K | $0.26 / $0.38 | ||
| 30 | DeepSeek | 685B | — | — | ||
| 33 | OpenAI | — | 400K | $0.20 / $1.25 | ||
| 34 | Alibaba Cloud / Qwen Team | 27B | 262K | $0.30 / $2.40 | ||
| 35 | Zhipu AI | 358B | 205K | $0.60 / $2.20 | ||
| 36 | Alibaba Cloud / Qwen Team | 35B | 262K | $0.25 / $2.00 | ||
| 37 | Xiaomi | 309B | — | — | ||
| 38 | Alibaba Cloud / Qwen Team | 480B | — | — | ||
| 39 | 120B | — | — |
FAQ
Common questions about Terminal-Bench 2.0.
More evaluations to explore
Related benchmarks in the same category
A challenging dataset of 448 multiple-choice questions written by domain experts in biology, physics, and chemistry. Questions are Google-proof and extremely difficult, with PhD experts reaching 65% accuracy.
A more robust and challenging multi-task language understanding benchmark that extends MMLU by expanding multiple-choice options from 4 to 10, eliminating trivial questions, and focusing on reasoning-intensive tasks. Features over 12,000 curated questions across 14 domains and causes a 16-33% accuracy drop compared to original MMLU.
All 30 problems from the 2025 American Invitational Mathematics Examination (AIME I and AIME II), testing olympiad-level mathematical reasoning with integer answers from 000-999. Used as an AI benchmark to evaluate large language models' ability to solve complex mathematical problems requiring multi-step logical deductions and structured symbolic reasoning.
Massive Multitask Language Understanding benchmark testing knowledge across 57 diverse subjects including STEM, humanities, social sciences, and professional domains
A verified subset of 500 software engineering problems from real GitHub issues, validated by human annotators for evaluating language models' ability to resolve real-world coding issues by generating patches for Python codebases.
Humanity's Last Exam (HLE) is a multi-modal academic benchmark with 2,500 questions across mathematics, humanities, and natural sciences, designed to test LLM capabilities at the frontier of human knowledge with unambiguous, verifiable solutions