t2-bench
t2-bench is a benchmark for evaluating agentic tool use capabilities, measuring how well models can select, sequence, and utilize tools to solve complex tasks. It tests autonomous planning and execution in multi-step scenarios.
Gemini 3.1 Pro from Google currently leads the t2-bench leaderboard with a score of 0.993 across 22 evaluated AI models.
Gemini 3.1 Pro leads with 99.3%, followed by Gemini 3 Flash at 90.2% and 
GLM-5 at 89.7%.
Progress Over Time
Interactive timeline showing model performance evolution on t2-bench
t2-bench Leaderboard
| Context | Cost | License | ||||
|---|---|---|---|---|---|---|
| 1 | Google | — | 1.0M | $2.50 / $15.00 | ||
| 2 | Google | — | 1.0M | $0.50 / $3.00 | ||
| 3 | Zhipu AI | 744B | 200K | $1.00 / $3.20 | ||
| 4 |  Alibaba Cloud / Qwen Team | 397B | 262K | $0.60 / $3.60 | ||
| 5 | Google | 31B | 262K | $0.14 / $0.40 | ||
| 6 | Google | 25B | 262K | $0.13 / $0.40 | ||
| 7 | Google | — | — | — | ||
| 8 | Alibaba Cloud / Qwen Team | 35B | 262K | $0.25 / $2.00 | ||
| 9 |  DeepSeek | 685B | 164K | $0.26 / $0.38 | ||
| 9 | DeepSeek | 685B | — | — | ||
| 11 | DeepSeek | 685B | — | — | ||
| 12 | Alibaba Cloud / Qwen Team | 4B | — | — | ||
| 13 | Alibaba Cloud / Qwen Team | 122B | 262K | $0.40 / $3.20 | ||
| 14 | Alibaba Cloud / Qwen Team | 9B | — | — | ||
| 15 | Alibaba Cloud / Qwen Team | 27B | 262K | $0.30 / $2.40 | ||
| 16 | Alibaba Cloud / Qwen Team | 1.0T | 256K | $0.50 / $5.00 | ||
| 17 |  LG AI Research | 236B | — | — | ||
| 18 |  OpenAI | 117B | 131K | $0.10 / $0.50 | ||
| 19 | Google | 8B | — | — | ||
| 20 | Alibaba Cloud / Qwen Team | 2B | — | — | ||
| 21 | Google | 5B | — | — | ||
| 22 | Alibaba Cloud / Qwen Team | 800M | — | — |
FAQ
Common questions about t2-bench.
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