τ-bench Banking benchmark

117 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

GPT-5.6 Sol (OpenAI) leads at 33%, from $5.00 per 1M input.

Data via GitHub: sierra-research/tau2-bench

Models ranked

117

on this benchmark

Top score

33%

current leader

Type

Single eval

benchmark

Updated

15 July 2026

last source fetch

Cost vs τ-bench Banking

73 priced models
Most attractive quadrant (cheap + high)
= reasoning model (extended thinking)

Each 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 · 117 models
Maker
#
1GPT-5.6 SolOpenAI33%$5.00$30.00
2Grok 4.5xAI33%$2.00$6.00
3GPT-5.6 TerraOpenAI32%$2.50$15.00
4GPT-5.5OpenAI31%$5.00$30.00
5Claude Sonnet 4.6Anthropic31%reasoning: adaptive$3.00$15.00
6GPT-5.4OpenAI30%$2.50$15.00
7Claude Opus 4.7Anthropic29%$5.00$25.00
8Claude Sonnet 5Anthropic28%$2.00$10.00
9JT-4.1 Flash 236B A21BChina-mobile28%n/an/a
10Claude Opus 4.8Anthropic28%$5.00$25.00
11GPT-5.6 LunaOpenAI27%$1.00$6.00
12Claude Fable 5Anthropic27%$10.00$50.00
13GLM 5.2Zhipu27%$1.40$4.40
14DeepSeek V4 ProDeepSeek26%$0.43$0.87
15Gemini 3.5 FlashGoogle25%$1.50$9.00
16Muse Spark 1.1Other25%$1.25$4.25
17DeepSeek V4 FlashDeepSeek23%n/an/a
18GPT-5.4 MiniOpenAI21%$0.75$4.50
19GPT-5.4 NanoOpenAI21%$0.20$1.25
20Hy3-previewTencent21%n/an/a
21Kimi K2.6Moonshot21%n/an/a
22GPT-5OpenAI20%$1.25$10.00
23Muse SparkMeta20%n/an/a
24Claude 4.5 SonnetAnthropic19%reasoning: true$3.00$15.00
25DeepSeek V3.2DeepSeek19%reasoning: true$0.27$0.40
26Kimi K2.7 CodeMoonshot18%$0.95$4.00
27Qwen3.7 PlusAlibaba18%n/an/a
28Gemini 3 FlashGoogle18%reasoning: truen/an/a
29Nex-N2-ProNex-agi17%n/an/a
30Gemini 3.1 Pro PreviewGoogle16%$2.00$12.00
31Qwen3.6 PlusAlibaba16%$0.33$1.95
32DeepSeek V3.1 TerminusDeepSeek16%reasoning: true$0.27$1.00
33Qwen3.6 27BAlibaba15%$0.15$0.50
34Gemma 4 31BGoogle15%$0.14$0.40
35GPT-5 MiniOpenAI15%$0.25$2.00
36Mistral Medium 3.5Mistral14%$1.50$7.50
37Kimi K2.5Moonshot14%$0.50$2.80
38Ring-2.6-1TInclusionai14%n/an/a
39GPT-5.1OpenAI14%$1.25$10.00
40Nemotron 3 UltraNVIDIA14%n/an/a
41Claude 4 SonnetAnthropic14%reasoning: true$3.00$15.00
42Qwen3.5-122B-A10BAlibaba14%$0.25$1.75
43Qwen3.5 397B A17BAlibaba13%$0.60$3.60
44MiniMax M3MiniMax13%$0.30$1.20
45LongCat 2.0Longcat13%n/an/a
46Grok 4.3xAI12%$1.25$2.50
47gpt-oss-120bOpenAI12%$0.05$0.25
48GPT-5.2OpenAI12%reasoning: false$1.75$14.00
49Gemma 4 26B A4bGoogle12%$0.13$0.40
50GLM 5.1Zhipu12%$1.05$3.50
150 of 117
Page 1 of 3
Prices are on-demand list rates. Enterprise commitments can be materially lower. Cost per task uses best-effort token usage.

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?

GPT-5.6 Sol (OpenAI) leads with 33%, from $5.00 per 1M input tokens.

How many models are ranked on τ-bench Banking?

117 models carry a τ-bench Banking score. Scores come from Artificial Analysis; where a model has a live API price it is verified against LiteLLM and OpenRouter, and models with no current price are listed without one.

Other benchmarks

compare the same models on a different eval
Benchmark scores via Artificial Analysis · pricing verified against LiteLLM + OpenRouter · refreshed daily · updated 15 July 2026

Every weekday

AI moves fast. Here's your debrief.

News, analysis, tools, and more.

For people who build with AI