MiniMax-M3 vs MiniMax-M3.1-Flash
Price and specs side by side · USD per 1M tokens · verified 2026-10-23
// side by side
| metric | MiniMax-M3 | MiniMax-M3.1-Flash |
|---|
| input / 1M | $0.30 | $0.15 |
| output / 1M | $1.20 | $0.60 |
| blended (3:1) | $0.525 | $0.2625 |
| cache read / 1M | $0.06 | $0.015 |
| context | 1M | 1M |
| max output | 524.3K | 524.3K |
| input types | text · image · video | text · image · video |
| reasoning | yes | yes |
| tool calling | yes | yes |
| structured output | no | no |
| open weights | yes | no |
| released | 2026-06-01 | 2026-10-19 |
| knowledge cutoff | n/a | n/a |
| lab | MiniMax | MiniMax |
// monthly cost scenarios
| workload | tokens in / out | MiniMax-M3 | MiniMax-M3.1-Flash |
|---|
| Chatbot | 10M / 2M | $5.40 | $2.70 |
| RAG app | 50M / 5M | $21.00 | $10.50 |
| Coding agent | 20M / 10M | $18.00 | $9.00 |
// verdict
- MiniMax-M3.1-Flash is 50% cheaper than MiniMax-M3 on a blended 3:1 input/output basis.
- Chatbot (10M input / 2M output tokens per month): MiniMax-M3.1-Flash costs $2.70 vs $5.40 for MiniMax-M3, saving $2.70 per month.
- RAG app (50M input / 5M output tokens per month): MiniMax-M3.1-Flash costs $10.50 vs $21.00 for MiniMax-M3, saving $10.50 per month.
- Coding agent (20M input / 10M output tokens per month): MiniMax-M3.1-Flash costs $9.00 vs $18.00 for MiniMax-M3, saving $9.00 per month.
// MiniMax-M3 or MiniMax-M3.1-Flash: which one to choose
- Choose MiniMax-M3.1-Flash for cost-sensitive, high-volume workloads.
- Choose MiniMax-M3 if you want open weights you can self-host.