AI models

τ-bench Verified (airline): AI model scores

130 AI models on τ-bench Verified (airline), per OpenRouter Benchmarks. Low discrimination: most models tie with the top score.

τ-bench Verified (airline), on the OpenRouter Benchmarks leaderboard (CC BY 4.0, latest version we checked), simulates an airline support agent that uses systems and must follow company policy until the end of the conversation. The score is the share of tasks completed within the rules. Here are the 130 models in this section with a score on τ-bench Verified (airline), per OpenRouter Benchmarks. Most are in a statistical tie with the top score, so the table shows the score without a rank or a badge.

τ-bench Verified (airline)

82%

Claude Fable 5

models

130

01

Model%Source
Claude Fable 5Anthropic · closed82% OpenRouter Benchmarks
DeepSeek V4 Flash 0423DeepSeek · open81.7% OpenRouter Benchmarks
Gemini 3.7 FlashGoogle · closed80.7% OpenRouter Benchmarks
Qwen3.5 397B A17BAlibaba · open80.2% OpenRouter Benchmarks
GPT-5.1OpenAI · closed80% OpenRouter Benchmarks
Ling 3.1 FlashAnt Group · undisclosed79.3% OpenRouter Benchmarks
Claude Fable 5.1Anthropic · closed79.2% OpenRouter Benchmarks
Qwen3.6 27BAlibaba · open79% OpenRouter Benchmarks
MiMo V2.6 ProXiaomi · open78.9% OpenRouter Benchmarks
Claude Sonnet 5.5Anthropic · closed78.8% OpenRouter Benchmarks
GLM-5.3Zhipu AI (Z.ai) · open78.8% OpenRouter Benchmarks
Claude Opus 5Anthropic · closed78.7% OpenRouter Benchmarks
Gemini 4 ArgonGoogle · closed78.7% OpenRouter Benchmarks
Kimi K3.1 ProMoonshot AI · open78.7% OpenRouter Benchmarks
Qwen3.5-35B-A3BAlibaba · open78.7% OpenRouter Benchmarks
DeepSeek V4 ProDeepSeek · open78.6% OpenRouter Benchmarks
Gemini 3.5 Flash LiteGoogle · closed78.3% OpenRouter Benchmarks
MiMo-V2.6-FlashXiaomi · open78.2% OpenRouter Benchmarks
Qwen3.8 27BAlibaba · open78% OpenRouter Benchmarks
Claude Opus 4.7Anthropic · closed77.8% OpenRouter Benchmarks
DeepSeek V4.1 FlashDeepSeek · open77.8% OpenRouter Benchmarks
Qwen3.5-122B-A10BAlibaba · open77.7% OpenRouter Benchmarks
Hy4Tencent · open77.6% OpenRouter Benchmarks
Claude Opus 4.8Anthropic · closed77.5% OpenRouter Benchmarks
Qwen3.8 2.4T A95BAlibaba · open77.4% OpenRouter Benchmarks
Kimi K2.7 CodeMoonshot AI · open77.3% OpenRouter Benchmarks
Muse Glimmer 30BMeta · open77.2% OpenRouter Benchmarks
Gemini 3.1 ProGoogle · closed77% OpenRouter Benchmarks
Gemma 4 31BGoogle · open77% OpenRouter Benchmarks
GLM 5Zhipu AI (Z.ai) · open76.9% OpenRouter Benchmarks
DeepSeek V4 Flash Vision ExpDeepSeek · open76.7% OpenRouter Benchmarks
Gemini 3.8 FlashGoogle · closed76.7% OpenRouter Benchmarks
GLM-5.2Zhipu AI (Z.ai) · open76.7% OpenRouter Benchmarks
Nova Micro 1.0Amazon · undisclosed76.7% OpenRouter Benchmarks
Claude Opus 4.5Anthropic · closed76.6% OpenRouter Benchmarks
Qwen3.8 FlashAlibaba · open76.4% OpenRouter Benchmarks
Gemini 3.6 FlashGoogle · closed76.3% OpenRouter Benchmarks
GLM-5.3-FlashZhipu AI (Z.ai) · open76.3% OpenRouter Benchmarks
MiniMax M3MiniMax · open76.3% OpenRouter Benchmarks
Kimi K3Moonshot AI · open76.2% OpenRouter Benchmarks
Qwen3.6 35B A3BAlibaba · open76.2% OpenRouter Benchmarks
GLM 4.7Zhipu AI (Z.ai) · open76% OpenRouter Benchmarks
Claude Opus 4.6Anthropic · closed75.8% OpenRouter Benchmarks
GLM 5.1Zhipu AI (Z.ai) · open75.8% OpenRouter Benchmarks
Muse Spark OpenMeta · open75.7% OpenRouter Benchmarks
Qwen3.5-9BAlibaba · open75.7% OpenRouter Benchmarks
Claude Sonnet 5Anthropic · closed75.4% OpenRouter Benchmarks
Claude Opus 5.5Anthropic · closed75.3% OpenRouter Benchmarks
GPT-5.3-CodexOpenAI · closed75.3% OpenRouter Benchmarks
Grok 4.6xAI · closed75.3% OpenRouter Benchmarks
Grok 4.7xAI · closed75.3% OpenRouter Benchmarks
GPT-5.4OpenAI · closed75% OpenRouter Benchmarks
Claude Sonnet 4.5Anthropic · closed74.9% OpenRouter Benchmarks
Kimi K2.6Moonshot AI · open74.9% OpenRouter Benchmarks
Nemotron 3.5 Super VLNVIDIA · open74.9% OpenRouter Benchmarks
Claude Sonnet 4.6Anthropic · closed74.8% OpenRouter Benchmarks
DeepSeek V3.2DeepSeek · open74.7% OpenRouter Benchmarks
GPT-5.5OpenAI · closed74.7% OpenRouter Benchmarks
MiMo-V2.5Xiaomi · open74.7% OpenRouter Benchmarks
MiniMax M3.1 FlashMiniMax · open74.7% OpenRouter Benchmarks
GPT-5.6 SolOpenAI · closed74.2% OpenRouter Benchmarks
MiMo-V2.5-ProXiaomi · open74.1% OpenRouter Benchmarks
Claude Haiku 5.5Anthropic · closed74% OpenRouter Benchmarks
Ling 3.0 FlashAnt Group · open74% OpenRouter Benchmarks
Kimi K2.5Moonshot AI · open73.8% OpenRouter Benchmarks
MiniMax M2.7MiniMax · open73.4% OpenRouter Benchmarks
Gemini 3 Flash PreviewGoogle · undisclosed73.3% OpenRouter Benchmarks
GLM 4.6Zhipu AI (Z.ai) · open73.3% OpenRouter Benchmarks
DeepSeek V3.1DeepSeek · open73.1% OpenRouter Benchmarks
DeepSeek V3.1 TerminusDeepSeek · open73% OpenRouter Benchmarks
Gemini 3.5 FlashGoogle · closed73% OpenRouter Benchmarks
Gemini 3.1 Flash LiteGoogle · closed72.7% OpenRouter Benchmarks
GPT-5.6 LunaOpenAI · closed72.7% OpenRouter Benchmarks
GPT-6 Astra ProOpenAI · undisclosed72.7% OpenRouter Benchmarks
Qwen3.7 PlusAlibaba · closed72.7% OpenRouter Benchmarks
GPT-5 MiniOpenAI · closed72.4% OpenRouter Benchmarks
GPT-5.6 TerraOpenAI · closed72.3% OpenRouter Benchmarks
Ling 3.0 Flash VLAnt Group · open72% OpenRouter Benchmarks
Qwen3.7 MaxAlibaba · closed72% OpenRouter Benchmarks
GPT-5.6 Sol ProOpenAI · undisclosed71.7% OpenRouter Benchmarks
Nemotron 3.5 LightningNVIDIA · open71.5% OpenRouter Benchmarks
GPT-5OpenAI · closed70.7% OpenRouter Benchmarks
GPT-5.6 Luna ProOpenAI · undisclosed70.7% OpenRouter Benchmarks
GPT-6.1 SolOpenAI · closed70.7% OpenRouter Benchmarks
Gemma 4 26B A4BGoogle · open70.6% OpenRouter Benchmarks
GPT-6 SolOpenAI · closed70.2% OpenRouter Benchmarks
GPT-6.1 TerraOpenAI · closed70% OpenRouter Benchmarks
GLM 4.7 FlashZhipu AI (Z.ai) · open69.7% OpenRouter Benchmarks
GPT-6 AstraOpenAI · closed69.7% OpenRouter Benchmarks
GPT-6 Luna ProOpenAI · undisclosed69.3% OpenRouter Benchmarks
MiniMax M2.5MiniMax · open69.2% OpenRouter Benchmarks
GPT-6 LunaOpenAI · closed68.4% OpenRouter Benchmarks
gpt-oss-120bOpenAI · open67.3% OpenRouter Benchmarks
GPT-5.4 MiniOpenAI · closed66.7% OpenRouter Benchmarks
Claude Haiku 4.5Anthropic · closed66.4% OpenRouter Benchmarks
GPT-5.4 NanoOpenAI · closed65.6% OpenRouter Benchmarks
Granite 4.2 8BIBM · open64% OpenRouter Benchmarks
Qwen3 235B A22B Thinking 2507Alibaba · open61.7% OpenRouter Benchmarks
GLM 4.6VZhipu AI (Z.ai) · open60.7% OpenRouter Benchmarks
Qwen3 Coder NextAlibaba · open60.7% OpenRouter Benchmarks
Gemini 2.5 ProGoogle · closed60% OpenRouter Benchmarks
Gemini 2.5 FlashGoogle · closed59.1% OpenRouter Benchmarks
DeepSeek R1DeepSeek · open58% OpenRouter Benchmarks
Qwen3 32BAlibaba · open54.3% OpenRouter Benchmarks
Gemini 2.5 Flash LiteGoogle · undisclosed53.7% OpenRouter Benchmarks
gpt-oss-20bOpenAI · open53.5% OpenRouter Benchmarks
GPT-4.1OpenAI · closed52.3% OpenRouter Benchmarks
Nemotron 3 Nano 30B A3BNVIDIA · open52% OpenRouter Benchmarks
GPT-5 NanoOpenAI · closed48.3% OpenRouter Benchmarks
Gemma 3 12BGoogle · open48% OpenRouter Benchmarks
Qwen3 VL 235B A22B InstructAlibaba · open48% OpenRouter Benchmarks
Qwen3-CoderAlibaba · open48% OpenRouter Benchmarks
Qwen3-235B-A22BAlibaba · open47.9% OpenRouter Benchmarks
Qwen3 Next 80B A3B InstructAlibaba · open47.5% OpenRouter Benchmarks
Qwen3 30B A3BAlibaba · open47% OpenRouter Benchmarks
Mistral Small 4Mistral AI · open46% OpenRouter Benchmarks
Qwen3 Coder 30B A3B InstructAlibaba · open46% OpenRouter Benchmarks
GPT-4.1 MiniOpenAI · closed45.1% OpenRouter Benchmarks
Qwen3 14BAlibaba · open44% OpenRouter Benchmarks
DeepSeek V3DeepSeek · open43.8% OpenRouter Benchmarks
Llama 4 MaverickMeta · open43.4% OpenRouter Benchmarks
GPT-4oOpenAI · closed43.2% OpenRouter Benchmarks
Qwen3 30B A3B Instruct 2507Alibaba · open41.5% OpenRouter Benchmarks
Qwen3 VL 30B A3B InstructAlibaba · open41.3% OpenRouter Benchmarks
Llama 3.3 70B InstructMeta · open39.9% OpenRouter Benchmarks
Qwen3 VL 8B InstructAlibaba · open32.7% OpenRouter Benchmarks
GPT-4o-miniOpenAI · closed30.7% OpenRouter Benchmarks
Qwen2.5 72B InstructAlibaba · open27.3% OpenRouter Benchmarks
Llama 3.1 8B InstructMeta · open23.5% OpenRouter Benchmarks
GPT-4.1 NanoOpenAI · closed14% OpenRouter Benchmarks

FAQ

Which model has the top score on τ-bench Verified (airline)?

The top score on τ-bench Verified (airline) belongs to Claude Fable 5, at 82% per OpenRouter Benchmarks. Most models are in a statistical tie with it, so the table shows the score without a rank or a badge.

What does τ-bench Verified (airline) measure?

τ-bench Verified (airline), on the OpenRouter Benchmarks leaderboard (CC BY 4.0, latest version we checked), simulates an airline support agent that uses systems and must follow company policy until the end of the conversation. The score is the share of tasks completed within the rules.

Which models are in this ranking?

The 130 models in this section with a published score on τ-bench Verified (airline), per OpenRouter Benchmarks. No version inherits another's score.

How to read

  1. Benchmark scores are percentages, and higher is better.
  2. Rank is shown as a range, «1st of 130». Within the range, the gap to neighboring models falls within the benchmark's margin. The «top score» badge marks the model with the benchmark's top score, and the «statistical tie with the top score» badge marks models whose gap to it is within the margin.
  3. On a saturated benchmark, the table shows the scores with the badge and without a rank order.

Sources

  • Epoch AI, CC BY 4.0, latest version we checked.
  • OpenRouter Benchmarks, CC BY 4.0, latest version we checked.
  • OpenRouter, dados públicos da API; exports de benchmarks e rankings em CC BY 4.0, latest version we checked.

Dates come from the sources: OpenRouter listing or Hugging Face repo creation, which may differ from the official launch. Family and version were grouped by rule from the name.

How to cite

If you use this data, cite the page and the date you accessed it.

Adriano Meirinho. "τ-bench Verified (airline): AI model scores". meirinho.com.br. https://meirinho.com.br/en/ia/modelos/benchmark/tau-bench-airline/. Accessed November 21, 2026.

Third-party data follows the license of each source. The full table is available as CSV and JSON.

Markdown