RobustBench-TC - Query 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 a paraphrased user query; higher is better.

Metric: Tool-call accuracy (%) under the QueryPara perturbation (observation component: user query (semantic rephrasing)); 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-32B76.4
2Qwen 3 32BNon-reasoning74.9#983
3Qwen 3 14BNon-reasoning72.9#1284
4LoopTool-8B68.8#1497
5Qwen 3.5 9BNon-reasoning67.8#838
6Qwen 3 8BNon-reasoning66.8#1890
7ToolRL-Qwen2.5-1.5B66.8#2330
8MUA-RL-32B65.8
9MUA-RL-8B65.3#1656
10ToolRL-Qwen2.5-3B64.3
11TL-CodeLLaMA-262.8
12O4 Mini62.3#372
13MUA-RL-14B62.3#1674
14DeepSeek R1 Distill Qwen 14B61.8#2739
15ToolRL-Llama3.2-3B59.8

No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=robustbench-tc-query-paraphrase · How It Works · Data refreshed daily, snapshot 2026-11-22.