| # | model | served by | τ²-bench | share of leader | in / out per M |
|---|---|---|---|---|---|
| 01 |  minimax-m3 | MiniMax | 88.9% | $0.30 / $1.20 | |
| 02 |  nvidia-nemotron-3-ultra | Doubleword | 83.3% | $0.50 / $2.50 |
Bars are each score's share of the leader's, so a short bar is a real gap and not a rounding difference. One row per model family, using the lab's own endpoint where it exists.
minimax-m3 leads nvidia-nemotron-3-ultra by 5.6 points
not enough entries to show a distribution
Best for tool use 2 ranked by τ²-bench
τ²-Bench measures multi-turn agentic tool use: calling functions, following policies, and completing realistic tasks over many turns. If you are building agents or tool-calling workflows, this predicts real-world reliability better than single-shot benchmarks.
method
Scores for τ²-Bench come from Artificial Analysis, an independent benchmarking service. When a model is served by several providers (Anthropic direct, AWS Bedrock, Google Vertex), one canonical entry represents the model family so the ranking is not padded with duplicates. Benchmarks measure specific skills: validate on your own workload before committing.
Built from the same catalog the router reads at request time, rebuilt daily.
other lists 8
one api for every model on this list
Requesty is OpenAI-compatible. Switch between any two models above by changing one parameter, and keep the failover, caching and spend controls.