MiniMax
MiniMax ranks #11 of 91 providers on BenchLeader by its best model, with 7 ranked models. Its best model is MiniMax M3.1 Flash at 61.9 on the BenchLeader Index, strongest in reasoning. Prices run from $0.262 to $0.963 per million tokens blended; MiniMax M3.1 Flash is the cheapest ranked option. The fastest is MiniMax-M2 at 99 tokens per second. 10 of its 12 models ship open weights. It has shipped 8 dated releases, roughly one every 57 days; the last was 19 Oct 2026, so the next would be due between 22 Nov 2026 and 6 Jan 2027 on that pattern.
Consumer-app-driven lab (Talkie, Hailuo video) that also ships the open-weights MiniMax M-series text models.
- Provider rank
- #11
- by best model
- Best index
- 61.9
- MiniMax M3.1 Flash
- Ranked models
- 7
- 10 open weights
- Cheapest ranked
- $0.262/M
- MiniMax M3.1 Flash
Release timeline
Each model’s best configuration by release date. The top eight are labelled; hover any point for its name.
- MiniMax7
Next release
Over the last two years MiniMax shipped 8 distinct models, one every 57 days on median. The latest arrived on 19 Oct 2026.
If the cadence holds, the next one lands between 22 Nov 2026 and 6 Jan 2027.
An estimate from past release dates only. It knows nothing about roadmaps or announcements.
Fastest ranked model: MiniMax-M2 at 99 tok/s.
Head to head
MiniMax vs Anthropic
MiniMax vs Google
MiniMax vs Meta
MiniMax vs OpenAI
MiniMax vs Moonshot AI
MiniMax vs SpaceXAI
MiniMax vs DeepSeek
MiniMax vs Xiaomi
MiniMax vs Zhipu AI
MiniMax vs Alibaba
MiniMax vs StepFun
MiniMax vs Sapiens AI
MiniMax vs ByteDance
MiniMax vs Motif Technologies
MiniMax vs Mistral AI
MiniMax vs MBZUAI
MiniMax vs Meituan
MiniMax vs Multiverse Computing
MiniMax vs NVIDIAMiniMax models compared
Models
Every MiniMax model with a result, one row per model; switch to all configurations to see each reasoning effort.
| # | ||||||||
|---|---|---|---|---|---|---|---|---|
| 1 | NewMiniMax M3.1 Flashopen ↗ | 61.9±6 | $0.262 | 96 | 22 s | 1M | 2026-10-19 | 2026-10-23 |
| 2 | MiniMax M3open ↗ | 56.4±5 | $0.525 | 89 | 24 s | 1M | 2026-06-01 | 2026-10-23 |
| 3 | MiniMax M2.7open ↗ | 55.6±4 | $0.525 | 55 | 47 s | 205k | 2026-03-18 | 2026-10-23 |
| 4 | MiniMax-M2.1open ↗ | 53.8±3 | $0.525 | 68 | 31 s | 205k | 2025-12-23 | 2026-10-23 |
| 5 | MiniMax M2.5open ↗ | 53.6±4 | $0.525 | 65 | 32 s | 205k | 2026-02-12 | 2026-10-23 |
| 6 | MiniMax-M2open ↗ | 50.8±5 | $0.525 | 99 | 22 s | 205k | 2025-10-27 | 2026-10-23 |
| 7 | MiniMax-M1open | 47.5±4 | $0.963 | 95 | 23 s | 1M | 2025-06-13 | 2026-10-22 |
| 8 | Minimax 2.1 | – | – | – | – | – | – | 2026-01-12 · stale |
| 9 | MiniMax-01open ↗ | – | $0.425 | 27 | 0.70 s | 1.0M | – | – |
| 10 | MiniMax M2-her | – | $0.525 | 47 | 0.74 s | 66k | – | – |
| 11 | MiniMax-M2.5-highspeedopen | – | $1.05 | – | – | 205k | 2026-02-13 | – |
| 12 | MiniMax-M2.7-highspeedopen | – | $1.05 | – | – | 205k | 2026-03-18 | – |
Latest news
- China's MiniMax completes export registration for open-source, overseas AI model licensingReuters · 23 Oct 2026
- MiniMax launches M3.1 Flash model, says it cuts AI agent costs by 70%SiliconANGLE · 19 Oct 2026
- DeepSeek, MiniMax Win Approval for Global Use, State Media SaysBloomberg.com · 23 Oct 2026
- MiniMax just halved its token prices. Your AI agent bill may not follow.The New Stack · 19 Oct 2026
- DeepSeek, MiniMax secure licences to deploy AI overseas under China export rulesSouth China Morning Post · 23 Oct 2026
- MiniMax M3.1 Flash Launches, Cuts AI Costs 70% [2026]shattered.io · 19 Oct 2026
- MiniMax M3.1 Flash shows China's AI price race has no floor in sightThe Decoder · 19 Oct 2026
- MiniMax's new M3.1 Flash beats DeepSeek and Opus 5 on a key business benchmark at lower cost — but with a high hallucination rateVentureBeat · 19 Oct 2026