Which model tops the best-for-agents list right now?
GPT-5.6 Sol (OpenAI), as of 9 September 2026, with LMArena Agent 7.7% · AA τ²-Bench Telecom 85.1% across 2 public agent boards and ModelCap Index position #5.
Agent ranking
Best AI models for agents
Language models measured on at least two public agent boards (LMArena Agent, BFCL, τ²-bench), ordered by the agent component of the ModelCap Index. Models with a single agent result follow in their own tier, ordered by that board, so one result never outranks two.
Snapshot as of 9 September 2026
GPT-5.6 Sol from OpenAI leads the best-for-agents list as of 9 September 2026 with LMArena Agent 7.7% · AA τ²-Bench Telecom 85.1% across 2 public agent boards, holding ModelCap Index position #5; 29 models qualify.
Models on this list
29
With openly downloadable weights
9
With a listed API price
29
Best AI models for agents (4)
Every row is a current, canonical language model with a public position on the ModelCap Index; the list filters by a published fact and orders by a published figure.
Measured on one agent board (25)
Ranked models with a single admitted agent result, ordered by that board's normalised score. One board is a narrower claim than two, so these rows sit apart from the list above and never mix into it.
Best AI models for agents: common questions
How is the best-for-agents list built?
It contains current, canonical language models with a measured public ModelCap Index position and admitted results on at least two public agent-family boards, ordered by the agent component of the ModelCap Index (ties broken by Index rank). A second tier, "Measured on one agent board", follows: Ranked models with a single admitted agent result, ordered by that board's normalised score. One board is a narrower claim than two, so these rows sit apart from the list above and never mix into it. Nothing is estimated for the list itself: the filter is a published fact and the order is a published figure, and every row links to the model page that shows its evidence.
How many models qualify for the best-for-agents list?
29 models qualify (4 in the primary tier, 25 measured on one board). The population is the live ModelCap Index, so a model appears here the moment it holds a public position and meets the filter.
Which model on this list ranks highest on the ModelCap Index?
GPT-6 Astra holds ModelCap Index position #1, the highest of any model in the list, and sits at #2 here by agent boards.
What is the cheapest model on the best-for-agents list?
Solar Pro 4 lists the lowest blended API price at $0.052 per 1M tokens ($0.03 input / $0.12 output).
Which is the best open-weight model on this list?
MiMo-V2.5-Pro from Xiaomi is the highest-placed model with openly downloadable weights (#4 here, ModelCap Index #21).
How current is the best-for-agents list?
It is rendered from the sealed dataset published 9 September 2026 and re-renders within a minute of each refresh; positions, prices and context figures are the ones shown on the live rankings at the same instant.
Explore further
- AI model rankingsThe full live ModelCap Index
- All ranking lists6 lists cut from the Index
- GPT-5.6 Sol vs GPT-5.5Head-to-head on benchmarks, price and context
- GPT-5.6 Sol vs Kimi K3Head-to-head on benchmarks, price and context
- GPT-5.5 vs Kimi K3Head-to-head on benchmarks, price and context
- LMArena Agent leaderboard37 models with a published result
- BFCL V4 function-calling leaderboard32 models with a published result
- τ²-Bench Telecom (Artificial Analysis run)136 models with a published result
- Compare AI modelsAny two ranked models side by side