Best-value AI models
The frontier contains models for which no eligible alternative is at least as capable and no more expensive, with a strict advantage in one of those measures. It helps you compare trade-offs between benchmark quality and cost.
Which should you choose?
No single benchmark covers every use case. Try your own prompts and compare accuracy, response time, input/output support, and the actual bill.
AI models
| Model | Intelligence score | Benchmark cost (USD) | Input price / 1M tokens (USD) | Output price / 1M tokens (USD) | Output speed |
|---|---|---|---|---|---|
| GPT-6 Luna (Low)OpenAIBest value frontier | 21.50 | $10.6 | 0.10 | 0.50 | 115 tok/s |
| GPT-6 Luna (Medium)OpenAIBest value frontier | 29.90 | $31.2 | 0.10 | 0.50 | 92 tok/s |
| GPT-6 Luna (High)OpenAIBest value frontier | 32.90 | $47.9 | 0.10 | 0.50 | 121 tok/s |
| Claude Haiku 5.5 (Medium)AnthropicBest value frontier | 34.50 | $56.4 | 0.10 | 0.50 | 179 tok/s |
| GPT-6 Luna (Xhigh)OpenAIBest value frontier | 34.60 | $66.8 | 0.10 | 0.50 | 127 tok/s |
| Claude Haiku 5.5 (High)AnthropicBest value frontier | 37.80 | $94.1 | 0.10 | 0.50 | 179 tok/s |
| MiMo-V2.6-FlashXiaomiBest value frontier | 37.90 | $110 | 0.14 | 0.28 | 56 tok/s |
| GPT-6 Luna (Max)OpenAIBest value frontier | 38.10 | $122 | 0.10 | 0.50 | 140 tok/s |
| Claude Haiku 5.5 (Xhigh)AnthropicBest value frontier | 41.20 | $158 | 0.10 | 0.50 | 185 tok/s |
| MiMo-V2.6-ProXiaomiBest value frontier | 46.30 | $207 | 0.43 | 0.87 | 55 tok/s |
| GPT-6.1 Sol (Medium)OpenAIBest value frontier | 47.80 | $361 | 2.00 | 10.00 | 48 tok/s |
| GPT-6.1 Sol (High)OpenAIBest value frontier | 50.20 | $521 | 2.00 | 10.00 | 52 tok/s |
| GPT-6.1 Sol (Xhigh)OpenAIBest value frontier | 51.00 | $662 | 2.00 | 10.00 | 51 tok/s |
| GPT-6.1 Sol (Max)OpenAIBest value frontier | 51.80 | $1082 | 2.00 | 10.00 | 57 tok/s |
| Claude Opus 5.5 (High, Default Fallback)AnthropicBest value frontier | 53.60 | $2172 | 4.00 | 20.00 | 78 tok/s |
| Kimi K3.1 ProKimiBest value frontier | 55.10 | $3117 | 2.00 | 10.00 | 63 tok/s |
| Claude Opus 5.5 (Xhigh, Default Fallback)AnthropicBest value frontier | 56.00 | $4057 | 4.00 | 20.00 | 82 tok/s |
| Claude Opus 5.5 (Max, Default Fallback)AnthropicBest value frontier | 57.60 | $8708 | 4.00 | 20.00 | 96 tok/s |
Guides
How we compare models
Intelligence uses the Artificial Analysis Intelligence Index. Benchmark cost is the total USD cost of running that test, including reasoning tokens; it is not a per-token API price. Rankings use the same dataset and calculations in every language. General benchmark scores do not measure performance in a particular language or task.