τ-bench Banking benchmark
148 AI models ranked on τ-bench Banking. Each row shows the model’s score next to its cheapest API price per 1M tokens where it has one, so you can weigh quality against cost.
Top model today
Qwen3.8 Max (Alibaba) leads at 51%, from $1.65 per 1M input via DeepInfra.
Data via GitHub: sierra-research/tau2-bench →Models ranked
148
on this benchmark
Top score
51%
current leader
Type
Single eval
benchmark
Updated
Oct 14, 2026
benchmark scores
Cost vs τ-bench Banking
116 priced modelsEach labelled name links to that model’s page. Curated to the cost-efficiency frontier, top scorers and cheapest so labels stay readable; the full field is in the leaderboard table below (and via “select all”). Reasoning models (extended “thinking” before answering) carry a ring around the dot.
Models ranked on τ-bench Banking
highest score first · 148 models| # | ||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Qwen3.8 Max |  Alibaba | 51% | $1.65 via DeepInfra | $4.95 via DeepInfra | |||||||||||||||||||||||||||||
| 2 | Grok 4.6 |  xAI | 51% | $2.00 via xAI | $6.00 via xAI | |||||||||||||||||||||||||||||
| 3 | Muse Spark 1.3 | Other | 51% | $1.25 via Azure | $4.25 via Azure | |||||||||||||||||||||||||||||
| 4 | GLM 5.3 |  Zhipu | 50% | $0.079 Routed quote | $3.08 via Sail | |||||||||||||||||||||||||||||
| 5 | Qwen3.8 2.4T A95B | Alibaba | 49% | $2.00 via Aihubmix | $6.00 via Aihubmix | |||||||||||||||||||||||||||||
| 6 | Qwen3.8 27B | Alibaba | 48% | $0.04 Routed quote | $1.49 via Cerebras | |||||||||||||||||||||||||||||
| 7 | Qwen3.8 Max (0902) | Alibaba | 48% | $2.00 Routed quote | $6.00 Routed quote | |||||||||||||||||||||||||||||
| 8 | GLM 5.3 Flash | Zhipu | 47% | $0.11 via Sail | $0.35 via Sail | |||||||||||||||||||||||||||||
| 9 | Claude Fable 5.1 |  Anthropic | 47% | $10.00 via Anthropic | $50.00 via Anthropic | |||||||||||||||||||||||||||||
| 10 | Kimi K3 |  Moonshot | 46% | $0.70 Routed quote | $12.50 via Sail | |||||||||||||||||||||||||||||
| 11 | Qwen3.8-Flash-Next | Alibaba | 45% | n/a | n/a | |||||||||||||||||||||||||||||
| 12 | Gemini 3.8 Flash |  Google | 45% | $0.75 via Google | $3.75 via Google | |||||||||||||||||||||||||||||
| 13 | GPT-5.6 Sol |  OpenAI | 44% | $2.00 Routed quote | $10.00 Routed quote | |||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||
| 14 | Claude Opus 5 | Anthropic | 42% | $5.00 via Anthropic | $25.00 via Anthropic | |||||||||||||||||||||||||||||
| 15 | Grok 4.5 | xAI | 42% | $2.00 via xAI | $6.00 via xAI | |||||||||||||||||||||||||||||
| 16 | GPT-6 Astra | OpenAI | 41% | $10.00 via OpenAI | $50.00 via OpenAI | |||||||||||||||||||||||||||||
| 17 | DeepSeek V4 Flash Vision (Reasoning, Max Effort) |  DeepSeek | 41% | n/a | n/a | |||||||||||||||||||||||||||||
| 18 | GPT-5.6 Terra | OpenAI | 40% | $2.00 via OpenAI | $12.00 via OpenAI | |||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||
| 19 | DeepSeek V4 Pro 0813 | DeepSeek | 40% | $0.66 Routed quote | $1.98 Routed quote | |||||||||||||||||||||||||||||
| 20 | GPT-5.4 | OpenAI | 40% | $2.50 via Databricks | $15.00 via OpenAI | |||||||||||||||||||||||||||||
| 21 | DeepSeek V4 Flash 0731 | DeepSeek | 39% | $0.016 Routed quote | $0.153 Routed quote | |||||||||||||||||||||||||||||
| 22 | GPT-5.5 | OpenAI | 39% | $5.00 via OpenAI | $30.00 via OpenAI | |||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||
| 23 | inclusionAI: Ling 3.0 Flash Fin | Other | 39% | $0.075 Routed quote | $0.22 Routed quote | |||||||||||||||||||||||||||||
| 24 | Claude Fable 5 | Anthropic | 38% | $10.00 via Anthropic | $50.00 via Anthropic | |||||||||||||||||||||||||||||
| 25 | Claude Sonnet 5 | Anthropic | 37% | $2.00 via Anthropic | $10.00 via Anthropic | |||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||
| 26 | Agnes 2.5 Pro Beta | Sapiens-ai | 36% | n/a | n/a | |||||||||||||||||||||||||||||
| 27 | Motif 3 | Motif-technologies | 35% | n/a | n/a | |||||||||||||||||||||||||||||
| 28 | Muse Spark 1.2 | Other | 35% | $1.25 via Meta | $4.25 via Meta | |||||||||||||||||||||||||||||
| 29 | GLM 5.2 | Zhipu | 35% | $0.05 Routed quote | $1.98 via Scx AI | |||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||
| 30 | Claude Opus 4.7 | Anthropic | 35% | $5.00 via Anthropic | $25.00 via Anthropic | |||||||||||||||||||||||||||||
| 31 | Claude Sonnet 4.6 | Anthropic | 34%reasoning: adaptive | $3.00 via Databricks | $15.00 via Anthropic | |||||||||||||||||||||||||||||
| 32 | inclusionAI: Ling 3.0 Flash VL | Other | 34% | $0.075 Routed quote | $0.22 Routed quote | |||||||||||||||||||||||||||||
| 33 | Claude Opus 4.8 | Anthropic | 34% | $5.00 via Anthropic | $25.00 via Anthropic | |||||||||||||||||||||||||||||
| 34 | K2 Horizon 375B A23B | Mbzuai-institute-of-foundation-models | 34% | n/a | n/a | |||||||||||||||||||||||||||||
| 35 | Gemini 3.7 Flash | Google | 33% | $0.75 via Google | $3.75 via Google | |||||||||||||||||||||||||||||
| 36 | Gemini 3.5 Flash | Google | 32% | $1.50 via Google | $9.00 via Google | |||||||||||||||||||||||||||||
| 37 | Muse Spark 1.1 | Other | 32% | $1.25 via Meta | $4.25 via Meta | |||||||||||||||||||||||||||||
| 38 | GPT-5.6 Luna | OpenAI | 31% | $0.20 via OpenAI | $1.20 via OpenAI | |||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||
| 39 | DeepSeek V4 Flash 0423 | DeepSeek | 31% | $0.09 via DeepInfra | $0.18 via DeepInfra | |||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||
| 40 | DeepSeek V4 Pro 0423 | DeepSeek | 30% | $0.435 via Tencent | $0.87 via Tencent | |||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||
| 41 | Gemini 3.6 Flash | Google | 30% | $0.75 via Google | $3.75 via Google | |||||||||||||||||||||||||||||
| 42 | Inkling | Thinking Machines | 29% | $0.95 via DeepInfra | $4.05 via Baseten | |||||||||||||||||||||||||||||
| 43 | Quasar 438B | Multiverse-computing | 29% | n/a | n/a | |||||||||||||||||||||||||||||
| 44 | GPT-5.4 Nano | OpenAI | 27% | $0.20 via Databricks | $1.25 via Databricks | |||||||||||||||||||||||||||||
| 45 | inclusionAI: Ling 3.0 Flash | Other | 27% | $0.021 Routed quote | $0.063 Routed quote | |||||||||||||||||||||||||||||
| 46 | GPT-5.4 Mini | OpenAI | 26% | $0.75 via Databricks | $4.50 via OpenAI | |||||||||||||||||||||||||||||
| 47 | Apodex 1.1 | Apodex | 25% | n/a | n/a | |||||||||||||||||||||||||||||
| 48 | Claude 4.5 Sonnet | Anthropic | 25%reasoning: true | $3.00 via Replicate | $15.00 via Replicate | |||||||||||||||||||||||||||||
| 49 | Muse Glimmer |  Meta | 24% | n/a | n/a | |||||||||||||||||||||||||||||
| 50 | Kimi K2.6 | Moonshot | 23% | $0.65 via Wandb | $3.40 via Novita | |||||||||||||||||||||||||||||
Frequently asked questions
What is the τ-bench Banking benchmark?
The banking domain in Sierra Research's tau-bench family: a knowledge-retrieval customer-service scenario where an agent answers banking questions using a retrieval pipeline rather than action APIs. Added in the τ³-bench update to the tau2-bench repository.
Which AI model scores highest on τ-bench Banking?
Qwen3.8 Max (Alibaba) leads with 51%, from $1.65 per 1M input tokens from DeepInfra.
How many models are ranked on τ-bench Banking?
148 models carry a τ-bench Banking score. Scores come from Artificial Analysis; where a model has a listed API price it is drawn from LiteLLM and OpenRouter, and models with no current price are listed without one.
Other benchmarks
compare the same models on a different evalEvery weekday
AI moves fast. Here's your debrief.
News, analysis, tools, and more.
For people who build with AI