Model intelligence
4 AI models
Capability (Epoch Capabilities Index and individual benchmarks from Epoch AI), API list prices and context windows (OpenRouter), release dates and open-weight status. Benchmarks are shown individually - never merged into one invented score.
Epoch Capabilities Index by release date
Capability frontier
Source: Epoch AI (CC BY 4.0). ECI is Epoch's item-response model over many benchmarks; each model page shows its 90% confidence interval.
Blended $/1M tokens (3:1 input:output) vs ECI
Price vs capability
Only models with both an ECI score and an OpenRouter price. Normalisation: 75% input + 25% output price. Click a point to open the model.
| # | Model | ECI ↑ | Context | In $/M | Weights | Released |
|---|---|---|---|---|---|---|
| 1 | Kimi K2.5 Moonshot AI· reasoning | 148.1 | 262K | $0.45 | Open source | 27 Jan 2026 |
| 2 | Kimi K2.7 Code Moonshot AI· reasoning | 150.0 | 262K | $0.66 | Open source | 12 Jun 2026 |
| 3 | Kimi K2.6 Moonshot AI· reasoning | 151.0 | 262K | $0.65 | Open source | 20 Apr 2026 |
| 4 | Kimi K3 Moonshot AI· reasoning | 157.7 | 1.05M | $3.00 | Open source | 16 Jul 2026 |
Showing 4 of 4. Benchmark cells are % (best reported setting). 3.8K benchmark results in total.Export CSV