Kimi K3 vs Qwen3.8 27B
API pricing, context, and independently measured performance, side by side. Qwen3.8 27B starts 94% cheaper on input than Kimi K3 at list prices.
Kimi K3
Moonshot AI · released Jul 16, 2026
Best input$1.70 / MTok
Best output$8.50 / MTok
Context1.05M
Providers20
Fastest measured0.63s° TTFT · 102 tok/s
Open weightsVisionToolsReasoningStructured outputPrompt caching
Qwen3.8 27B
Qwen · released Aug 14, 2026
Best input$0.10 / MTok
Best output$1.80 / MTok
Context1M
Providers17
Fastest measured0.91s° TTFT · 33 tok/s
Open weightsVisionToolsReasoningStructured outputPrompt caching
Cheapest providers, measuredUSD / MTOK
| MODEL | PROVIDER | INPUT | OUTPUT | TTFT | TPS | SMOKE |
|---|---|---|---|---|---|---|
| Kimi K3 | RERelace | $1.70 | $8.50 | — | — | — |
| Kimi K3 | INInferenceNet | $2.10 | $10.95 | 57.9s° | 7° tok/s | — |
| Kimi K3 | WAWafer | $2.10 | $10.95 | 29.3s° | 89° tok/s | 0/5 |
| Qwen3.8 27B | DADarkbloom | $0.10 | $1.80 | — | — | — |
| Qwen3.8 27B |  DeepInfra | $0.15 | $1.88 | — | — | — |
| Qwen3.8 27B | PHPhala | $0.199 | $2.08 | 0.91s° | 33° tok/s | — |
TTFT/TPS: ModelIndex-measured medians (P50, trailing 21d) · chat10k profile (~10K in / 200 out) · ° = measured via OpenRouter routing (adds a hop) · SMOKE: of 5 programmatic capability probes passed · methodology v1.1 — full details on each model page