MiniMax M3.1 Flash Joins the Flash Race
Matches DeepSeek’s off-peak pricing, halves peak rates, trails it on four of six benchmarks.
Another Chinese Flash model arrives days after DeepSeek’s, matching its off-peak price.
MiniMax has officially joined the “Flash party,” @MiaAI_lab says: MiniMax-M3.1-Flash is coming. The mystery model Space Bunny “was indeed MniMax,” as she had suspected.
@kimmonismus posted the official benchmarks and calls them “very impressive,” since DeepSeek’s V4.1 Flash is only a few days old and MiniMax’s new Flash model outperforms it in several key benchmarks.
| Measure | MiniMax M3.1 Flash vs. DeepSeek V4.1 Flash |
|---|---|
| Off-peak input/output price | Matches |
| Peak rates | Half |
| Cached input price | Higher |
| Comparable benchmarks | Trails on 4 of 6 |
For reference, Sunday’s edition listed DeepSeek V4.1 Flash-1017 at $0.15 / $0.60 per million uncached input/output tokens off-peak and $0.30 / $1.20 at peak.
@kimmonismus frames it as a race: “China has likely heard the call and now needs to catch up quickly, as the U.S. remains clearly in the lead.”
Sources
- @MiaAI_lab on X · Oct 19, 2:29 AM PT
- @kimmonismus on X · Oct 19, 4:57 AM PT
- @kimmonismus on X · Oct 19, 3:30 AM PT