| Spec | MiniMax M3 | DeepSeek V4 Pro |
|---|---|---|
| Lowest input | $0.28DeepInfra | $1.30DeepInfra |
| Lowest output | $1.10DeepInfra | $2.60DeepInfra |
| Lowest cached input | $0.056DeepInfra | $0.0036Inference.net |
| Context | 524K | 1.05M |
| Max output | 131K | 524K |
| Providers | 3 | 4 |
| Released | June 1, 2026 | April 24, 2026 |
| Weights | Open | Open |
| Lab | MiniMax | DeepSeek |
| Model ID | minimax/minimax-m3 | deepseek/deepseek-v4-pro |
USD per 1M tokens, as of November 5, 2026.
| Provider | MiniMax M3 | DeepSeek V4 Pro |
|---|---|---|
| DeepInfra | $0.28 / $1.10 | $1.30 / $2.60 |
| StreamLake | $0.60 / $2.40 | $1.74 / $3.48 |
| Inference.net | – | $2.10 / $4.20 |
| Nebius | $0.30 / $1.20 | – |
| Novita AI | – | $1.60 / $3.20 |
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 |
|---|---|---|---|
| MiniMax M3DeepInfra | $1.11 | $19.50 | $195.00 |
| DeepSeek V4 ProDeepInfra | $3.90 | $78.00 | $780.00 |
On each model's cheapest provider by input plus output list price.
Input: MiniMax M3 at $0.28 per million input tokens vs $1.30 (DeepSeek V4 Pro), 4.6x lower. Output: MiniMax M3 at $1.10 per million output tokens vs $2.60 (DeepSeek V4 Pro), 2.4x lower. Lowest listed prices as of November 5, 2026.
DeepSeek V4 Pro, with 1.05M tokens vs 524K.
2 providers serve both: DeepInfra, StreamLake. MiniMax M3 has 3 providers in total and DeepSeek V4 Pro has 4.
MiniMax M3 from $0.056 per million cached input tokens (DeepInfra). DeepSeek V4 Pro from $0.0036 (Inference.net).
Yes. Both use the same OpenAI-compatible endpoint, https://api.routerplus.com/v1, and key. Change the model field from "minimax/minimax-m3" to "deepseek/deepseek-v4-pro".
