Thinking Kimi model for slower research passes, planning, and hard technical questions
Benchmarks
| Benchmark | Category | Score | Variant | Provenance | Source | Observed |
|---|---|---|---|---|---|---|
| AA Agentic Index | agentic | 1.8 | thinking | mirror | openrouter | Jul 23, 2026 |
| AA Coding Index | coding | 21.0 | thinking | mirror | openrouter | Jul 23, 2026 |
| AA Intelligence Index | intelligence | 17.3 | thinking | mirror | openrouter | Jul 23, 2026 |
| Design Arena — codecategories | preference | 1131 | — | mirror | openrouter | Jul 23, 2026 |
| Design Arena — dataviz | preference | 1053 | — | mirror | openrouter | Jul 24, 2026 |
| Design Arena — gamedev | preference | 1032 | — | mirror | openrouter | Jul 24, 2026 |
| Design Arena — uicomponent | preference | 1072 | — | mirror | openrouter | Jul 23, 2026 |
| Design Arena — website | preference | 1140 | thinking | mirror | openrouter | Jul 23, 2026 |
| 1135 | — | mirror | openrouter | Jul 24, 2026 |
Includes index data from Artificial Analysis.
Providers
Live operational stats per hosting endpoint, via OpenRouter.
| Provider | Quant | Context | $/1M in · out | Uptime 24h | Latency | Throughput |
|---|---|---|---|---|---|---|
| — | 262K | $0.6 · $2.5 | 96.73% | — | — | |
| Groq | — | 262K | $1 · $3 | 99.94% | — | — |
| Novita | bf16 | 262K | $0.6 · $2.5 | 98.83% | — | — |
| Novita | fp8 | 131K | $0.57 · $2.3 | 99.98% | — | — |