Same open weights, served by 4 providers.
| Provider | Their model id | Serving | Input | Output | Cached in | Blended 3:1 | Source | |
|---|---|---|---|---|---|---|---|---|
DeepInfra Deprecated, 2026-09-10 | MiniMaxAI/MiniMax-M2.5 | standardFP8 | $0.15 | $1.15 | $0.03 | $0.40 | Verified 5h ago · Changed 7h ago | |
Together AI Deprecated | MiniMaxAI/MiniMax-M2.5 | standardQuantization not disclosed | $0.30 | $1.2 | $0.06 | $0.525 | Verified 7h ago · Changed 7h ago | |
AWS Bedrock | minimax.minimax-m2.5 | standardQuantization not disclosed | $0.30 | $1.2 | — | $0.525 | Verified 7h ago · Changed 7h ago | |
Azure AI Foundry Fireworks on Azure | minimax-m2.5 | standarddata zone (US)Quantization not disclosed | $0.33 | $1.32 | $0.033 | $0.578 | Verified 5h ago · Changed 7h ago | |
DeepInfra Deprecated, 2026-09-10 | MiniMaxAI/MiniMax-M2.5 | priorityFP81.5× standard | $0.225 | $1.73 | $0.045 | $0.60 | Verified 5h ago · Changed 7h ago |
Same weights don't always mean identical outputs: quantization, context limits and serving stacks can change quality, especially for tool calling. Prices are USD per 1M tokens; hyperscaler prices are for the reference region shown.
| DeepInfra fp8 | $38/mo |
| Together AI | $54/mo |
| AWS Bedrock | $54/mo |
| Azure AI Foundry data zone (US) | $59.4/mo |
Cache hits use each provider's cached-input price where it has one. Cache write costs aren't included.