Skip to main content

TAU3-Bench

A next-generation tool-use benchmark for complex, long-horizon agent workflows beyond the older tau2 telecom and airline task families.

Benchmark score on TAU3-Bench — May 13, 2026

BenchLM mirrors the published score view for TAU3-Bench. Mistral Medium 3.5 128B leads the public snapshot at 91.4% , followed by MiMo-V2.5-Pro (72.9%) and Qwen3.6 Plus (70.7%). BenchLM does not use these results to rank models overall.

9 modelsAgenticCurrentDisplay onlyUpdated May 13, 2026

The published TAU3-Bench snapshot is tightly clustered at the top: Mistral Medium 3.5 128B sits at 91.4%, while the third row is only 20.7 points behind. The broader top-10 spread is 25.8 points, so the benchmark still separates strong models even when the leaders cluster.

9 models have been evaluated on TAU3-Bench. The benchmark falls in the Agentic category. This category carries a 22% weight in BenchLM.ai's overall scoring system. TAU3-Bench is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.

About TAU3-Bench

Year

2026

Tasks

Long-horizon tool workflows

Format

Interactive tool-use evaluation

Difficulty

Advanced tool use

TAU3-Bench appears in the Qwen3.6 launch tables as a broader, more modern tool-use benchmark than the earlier tau2 slices. BenchLM tracks it separately because it is not the same benchmark as Tau2-Telecom or Tau2-Airline.

BenchLM freshness & provenance

Version

TAU3-Bench 2026

Refresh cadence

Quarterly

Staleness state

Current

Question availability

Public benchmark set

CurrentDisplay only

BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.

Benchmark score table (9 models)

1
91.4%
2
72.9%
3
70.7%
4
70.6%
5
70.2%
6
68.4%
7
67.2%
8
65.7%
9
65.6%

FAQ

What does TAU3-Bench measure?

A next-generation tool-use benchmark for complex, long-horizon agent workflows beyond the older tau2 telecom and airline task families.

Which model scores highest on TAU3-Bench?

Mistral Medium 3.5 128B by Mistral currently leads with a score of 91.4% on TAU3-Bench.

How many models are evaluated on TAU3-Bench?

9 AI models have been evaluated on TAU3-Bench on BenchLM.

Last updated: May 13, 2026 · BenchLM version TAU3-Bench 2026

The AI models change fast. We track them for you.

For engineers, researchers, and the plain curious — a weekly brief on new models, ranking shifts, and pricing changes.

Free. No spam. Unsubscribe anytime.