| Spec | DeepSeek V4 Pro | Kimi K3 |
|---|---|---|
| Lowest input | $1.30DeepInfra | $0.60Wafer |
| Lowest output | $2.60DeepInfra | $9.50Chutes |
| Lowest cached input | $0.0036Inference.net | $0.185Chutes |
| Context | 1.05M | 1.05M |
| Max output | 524K | 524K |
| Providers | 4 | 10 |
| Released | April 24, 2026 | July 16, 2026 |
| Weights | Open | Open |
| Lab | DeepSeek | Moonshot AI |
| Model ID | deepseek/deepseek-v4-pro | moonshotai/kimi-k3 |
USD per 1M tokens, as of November 5, 2026.
| Provider | DeepSeek V4 Pro | Kimi K3 |
|---|---|---|
| DeepInfra | $1.30 / $2.60 | $2.85 / $14.25 |
| Inference.net | $2.10 / $4.20 | $3.00 / $15.00 |
| Novita AI | $1.60 / $3.20 | $3.00 / $15.00 |
| Chutes | – | $1.85 / $9.50 |
| Doubleword | – | $3.00 / $15.00 |
| Morph | – | $2.50 / $14.00 |
| Nebius | – | $3.00 / $15.00 |
| Phala | – | $3.00 / $15.00 |
| Sail Research | – | $1.50 / $10.00 |
| StreamLake | $1.74 / $3.48 | – |
| Wafer | – | $0.60 / $15.00 |
Input / output, USD per 1M tokens.
| Model | 1,000 chat replies2K in, 500 out each | 1,000 agent tasks50K in, 5K out each | Monthly batch job500M in, 50M out |
|---|---|---|---|
| DeepSeek V4 ProDeepInfra | $3.90 | $78.00 | $780.00 |
| Kimi K3Chutes | $8.45 | $140.00 | $1,400.00 |
On each model's cheapest provider by input plus output list price.
Input: Kimi K3 at $0.60 per million input tokens vs $1.30 (DeepSeek V4 Pro), 2.2x lower. Output: DeepSeek V4 Pro at $2.60 per million output tokens vs $9.50 (Kimi K3), 3.7x lower. Lowest listed prices as of November 5, 2026.
Both list 1.05M tokens of context.
3 providers serve both: DeepInfra, Novita AI, Inference.net. DeepSeek V4 Pro has 4 providers in total and Kimi K3 has 10.
DeepSeek V4 Pro from $0.0036 per million cached input tokens (Inference.net). Kimi K3 from $0.185 (Chutes).
Yes. Both use the same OpenAI-compatible endpoint, https://api.routerplus.com/v1, and key. Change the model field from "deepseek/deepseek-v4-pro" to "moonshotai/kimi-k3".
