RobustBench-TC - Parameter Description Paraphrase: leaderboard

RobustBench-TC (2026) tests tool-calling models under deployment noise grounded in real tool-calling failures, using 199 single-turn samples drawn from BFCL, API-Bank, RoTBench, ToolAlpaca and ToolEyes and scored by deterministic rule-based parsers. This board is tool-call accuracy in percent under paraphrased parameter descriptions; higher is better.

Metric: Tool-call accuracy (%) under the ParamPara perturbation (observation component: parameter description); RobustBench-TC samples drawn from BFCL V3 single-turn, API-Bank, RoTBench, ToolAlpaca and ToolEyes, each scored by its source benchmark's strict scorer after a format-tolerant tool-call parse, temperature 0 (DeepSeek-R1-Distill at 0.6), Qwen3-family models with thinking disabled; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 18 models tracked. Updated 7 Oct 2026.

Top models

#ModelSettingScoreOverall rank
1LoopTool-32B77.9
2Qwen 3 32BNon-reasoning76.9#972
3Qwen 3 14BNon-reasoning73.9#1269
4ToolRL-Qwen2.5-1.5B70.4#2299
5LoopTool-8B68.8#1473
6Qwen 3.5 9BNon-reasoning68.3#828
7O4 Mini67.8#368
8MUA-RL-32B67.8
9Qwen 3 8BNon-reasoning67.3#1873
10MUA-RL-8B67.3#1640
11TL-CodeLLaMA-264.3
12MUA-RL-14B63.8#1658
13ToolRL-Llama3.2-3B63.8
14ToolRL-Qwen2.5-3B62.8
15DeepSeek R1 Distill Qwen 14B62.3#2771

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=robustbench-tc-parameter-description-paraphrase · How It Works · Data refreshed daily, snapshot 2026-11-21.