Wikibench

t2-bench

reasoning

Multilingual: no

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.

Methodology

Imported from llm-stats public benchmark metadata. Modality: text. Max score: 1. Categories: agents, reasoning, tool_calling. Language: en. Verified by llm-stats: no.

Leaderboard

Showing 17 of 17 results

  1. Gemini 3.1 Pro self-reported llm-stats
    99.3%
  2. Gemini 3 Flash self-reported llm-stats
    90.2%
  3. GLM-5 self-reported llm-stats
    89.7%
  4. Qwen3.5-397B-A17B self-reported llm-stats
    86.7%
  5. Gemma 4 31B self-reported llm-stats
    86.4%
  6. Gemma 4 26B-A4B self-reported llm-stats
    85.5%
  7. Gemini 3 Pro self-reported llm-stats
    85.4%
  8. Qwen3.5-35B-A3B self-reported llm-stats
    81.2%
  9. DeepSeek-V3.2 self-reported llm-stats
    80.3%
  10. DeepSeek-V3.2-Speciale self-reported llm-stats
    80.3%
  11. DeepSeek-V3.2 (Thinking) self-reported llm-stats
    80.2%
  12. Qwen3.5-122B-A10B self-reported llm-stats
    79.5%
  13. Qwen3.5-27B self-reported llm-stats
    79.0%
  14. K-EXAONE-236B-A23B self-reported llm-stats
    73.2%
  15. GPT OSS 120B High self-reported llm-stats
    63.9%
  16. Gemma 4 E4B self-reported llm-stats
    57.5%
  17. Gemma 4 E2B self-reported llm-stats
    29.4%