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 |
|---|---|---|---|---|---|
| Llama 4 ScoutMetaBest value frontier | 10.00 | $11.5 | 0.17 | 0.63 | 89 tok/s |
| DeepSeek V3 (Dec '24)DeepSeekBest value frontier | 14.20 | $19.8 | 0.36 | 0.89 | - |
| gpt-oss-120b (low)OpenAIBest value frontier | 14.90 | $24.2 | 0.15 | 0.60 | 315 tok/s |
| MiMo-V2.5XiaomiBest value frontier | 37.20 | $25.2 | 0.14 | 0.28 | 69 tok/s |
| DeepSeek V4 Flash (Reasoning, Max Effort)DeepSeekBest value frontier | 40.30 | $74.3 | 0.14 | 0.28 | 104 tok/s |
| MiMo-V2.5-ProXiaomiBest value frontier | 42.20 | $98.5 | 0.43 | 0.87 | 57 tok/s |
| DeepSeek V4 Pro (Reasoning, Max Effort)DeepSeekBest value frontier | 44.30 | $176 | 0.43 | 0.87 | 62 tok/s |
| MiniMax-M3MiniMaxBest value frontier | 44.40 | $204 | 0.30 | 1.20 | 84 tok/s |
| GPT-5.6 Terra (medium)OpenAIBest value frontier | 45.60 | $240 | 2.50 | 15.00 | 116 tok/s |
| GPT-5.6 Luna (high)OpenAIBest value frontier | 46.10 | $275 | 1.00 | 6.00 | 170 tok/s |
| GPT-5.6 Sol (low)OpenAIBest value frontier | 49.40 | $353 | 5.00 | 30.00 | 50 tok/s |
| Muse Spark 1.1 (xhigh)MetaBest value frontier | 50.60 | $548 | 1.25 | 4.25 | 118 tok/s |
| GPT-5.6 Sol (medium)OpenAIBest value frontier | 53.60 | $593 | 5.00 | 30.00 | 52 tok/s |
| Grok 4.5 (high)SpaceXAIBest value frontier | 53.80 | $602 | 2.00 | 6.00 | 103 tok/s |
| GPT-5.6 Sol (high)OpenAIBest value frontier | 55.90 | $956 | 5.00 | 30.00 | 47 tok/s |
| GPT-5.6 Sol (xhigh)OpenAIBest value frontier | 57.70 | $1543 | 5.00 | 30.00 | 52 tok/s |
| GPT-5.6 Sol (max)OpenAIBest value frontier | 58.90 | $2824 | 5.00 | 30.00 | 53 tok/s |
| Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback)AnthropicBest value frontier | 59.90 | $5631 | 10.00 | 50.00 | 66 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.