Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking
Benchmarks
| Benchmark | Category | Score | Variant | Provenance | Source | Observed |
|---|---|---|---|---|---|---|
| AA Agentic Index | agentic | 29.6 | — | mirror | openrouter | Jul 23, 2026 |
| AA Coding Index | coding | 60.8 | — | mirror | openrouter | Jul 23, 2026 |
| AA Intelligence Index | intelligence | 41.9 | — | mirror | openrouter | Jul 23, 2026 |
| Design Arena — 3d | preference | 1303 | — | mirror | openrouter | Jul 24, 2026 |
| Design Arena — agenticgamedev | preference | 1147 | — | mirror | openrouter | Jul 24, 2026 |
| Design Arena — androidnative | preference | 1261 | — | mirror | openrouter | Jul 24, 2026 |
| Design Arena — asciiart | preference | 1217 | — | mirror | openrouter | Jul 23, 2026 |
| Design Arena — codecategories | preference | 1291 | — | mirror | openrouter | Jul 23, 2026 |
| Design Arena — dataviz | preference | 1263 | — | mirror | openrouter | Jul 24, 2026 |
| Design Arena — fullstack | preference | 1239 | — | mirror | openrouter | Jul 24, 2026 |
| Design Arena — gamedev | preference | 1270 | — | mirror | openrouter | Jul 23, 2026 |
| Design Arena — godotgamedev | preference | 1209 | — | mirror | openrouter | Jul 24, 2026 |
| Design Arena — htmlslides | preference | 1220 | — | mirror | openrouter | Jul 24, 2026 |
| Design Arena — mobileapps | preference | 1220 | — | mirror | openrouter | Jul 23, 2026 |
| Design Arena — python-pptxslides | preference | 1151 | — | mirror | openrouter | Jul 23, 2026 |
| Design Arena — svg | preference | 1223 | — | mirror | openrouter | Jul 24, 2026 |
| Design Arena — uicomponent | preference | 1298 | — | mirror | openrouter | Jul 23, 2026 |
| Design Arena — webapps | preference | 1233 | — | mirror | openrouter | Jul 23, 2026 |
| Design Arena — website | preference | 1301 | — | mirror | openrouter | Jul 23, 2026 |
| LMArena WebDev | preference | 1470 | — | crowd | lmarena | Jul 21, 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 |
|---|---|---|---|---|---|---|
| Ambient | int4 | 262K | $0.71 · $3.49 | 99.75% | — | — |
| AtlasCloud | int4 | 262K | $0.95 · $4 | 98.79% | — | — |
| Cloudflare | — | 262K | $0.95 · $4 | 99.98% | — | — |
| CoreWeave | int4 | 262K | $0.94 · $4 | 99.77% | — | — |
| DeepInfra | fp4 | 262K | $0.74 · $3.5 | 98.08% | — | — |
| Fireworks | — | 262K | $0.95 · $4 | 98.82% | — | — |
| Inceptron | int4 | 262K | $0.72 · $3.5 | 99.85% | — | — |
| Moonshot AI | int4 | 262K | $1.9 · $8 | 99.82% | — | — |
| Moonshot AI | int4 | 262K | $0.95 · $4 | 99.98% | — | — |
| Novita | int4 | 262K | $0.912 · $3.84 | 99.76% | — | — |
| Parasail | int4 | 262K | $0.76 · $3.5 | 99.42% | — | — |
| SiliconFlow | fp8 | 262K | $0.859 · $3.8 | 99.91% | — | — |
| Together | — | 262K | $0.95 · $4 | 99.27% | — | — |
| Venice | int4 | 256K | $0.75 · $3.5 | 97.72% | — | — |