MiniMax M2
MiniMax M2Rankingin thinking mode?
Alternatives?
CheaperLing 3.0 Flash VLthinkingOpen×4.1-83%$0.06 / $0.18
BetterGPT-6 Lunalow+8%-62%$0.005Function calling only with none · Chat Completions+3OOpenMiniCPM5 1Bnon-thinking×100.9——Context: 205K → 131K
CheaperGranite 4.2 30BthinkingOpen×3.6-46%$0.072Context: 205K → 128KCheaperGranite 4.2 3BthinkingOpen×3.5-90%$0.006Context: 205K → 128KLing 3.0 Flash VLthinkingOpen×4.1-83%$0.06 / $0.18Granite 4.2 30BthinkingOpen×3.6-46%$0.072Context: 205K → 128KGranite 4.2 3BthinkingOpen×3.5-90%$0.006Context: 205K → 128K
GLM 5.3 FlashmaxOpen×3.3-55%$0.28Memory (4-bit): 137 GB → 193 GB+5
Gemma 4 E4BthinkingOpen×2.9-92%$0.008Context: 205K → 128KGPT-6 Lunalow+8%-62%$0.005Function calling only with none · Chat Completions+3Granite 4.2 3BthinkingOpen×3.5-90%$0.006Context: 205K → 128K
Ministral 3 3B 2512Open+13%-81%$0.008Context: 205K → 131KGemma 4 E4BthinkingOpen×2.9-92%$0.008Context: 205K → 128KGPT-5.6 Lunanon-thinking+21%-14%$0.01
Command A+thinkingOpen×6.4+14%$0.30 / $1.50Context: 205K → 128KlLFM2.5 2.6BthinkingOpen×5.7——Context: 205K → 131KMiniMax M3.1 Flashthinking-2%—$0.22Complete model response must be appendedMiniMax M1 80kOpensame—$0.55 / $2.20Memory (4-bit): 137 GB → 274 GBMiniMax M2.5thinkingOpen+3%—$0.30 / $1.20MiniMax M2.1thinkingOpen+33%—$0.30 / $1.20MiniMax M2.7thinkingOpen×2.6—$0.30 / $1.20MiniMax M3thinkingOpen×4.9—$0.51Memory (4-bit): 137 GB → 256 GB+2MiniMax M1 40kOpenn/a——Memory (4-bit): 137 GB → 274 GBMiniMax M1unstatedOpenn/a—$0.40 / $2.20Memory (4-bit): 137 GB → 274 GB+1All MiniMax models →Run it yourself
estimateHistory
6 versionsBenchmarks
all categoriesFAQ
How much does MiniMax M2 cost?›
$0.30 per 1M input tokens and $1.20 per 1M output tokens; cheapest via DeepInfra at $0.25 / $1.02.
Is MiniMax M2 free to use?›
There is no free API; the cheapest paid price is $0.25 per 1M input tokens. The weights are open, so you can run it yourself.
How good is MiniMax M2, and what is it best at?›
#152 of 334 in Chat & Knowledge, #86 of 97 in Coding, #54 of 242 in Agents & Tool Use, #142 of 195 in Writing & Language. Strongest at: Tool calling, Instructions.
Is MiniMax M2 better than Gemini 2.5 Pro?›
On AA Omniscience · Hallucination Rate, MiniMax M2 scores 90.9 vs 90.9 — about the same.
How can I use MiniMax M2?›
Through the API of AWS Bedrock, DeepInfra, Google.
What is the context window of MiniMax M2?›
205K tokens, with up to 131K output tokens.
When was MiniMax M2 released? Is there a newer version?›
Released 2025-10-27. It has been superseded by MiniMax M2.1.
Is MiniMax M2 open source?›
Yes, the weights are open (modified-mit licence).
Can I run MiniMax M2 locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
Can MiniMax M2 read images?›
No, it accepts text only.
What is MiniMax M2?›
MiniMax M2 is an open-weights language model from MiniMax, released Oct 2025, the successor to MiniMax M1 40k.
What are the best alternatives to MiniMax M2?›
Ling 3.0 Flash VL (cheaper), GPT-6 Luna (better), MiniCPM5 1B (open weights).