Compare models

Context, limits, capabilities, price and use cases side by side — up to 4 models at a time. The URL holds your selection, so you can share it.

MiniMax M2.7MiniMax-M3 Add a modelChat with these 2

AttributeHide identical rows	MiniMax M2.7minimax/minimax-m2.7	MiniMax-M3minimax/minimax-m3
Specs
Context window	205K	1M
Max output	131K	512K
Input modalities	Text	TextImageVideo
Output modalities	Text	Text
Capabilities
Streaming	Supported	Supported
Tool calling	Supported	Supported
Structured output (JSON)	Supported	Supported
Image input	Not supported	Supported
Thinking mode	Supported	Supported
Price / 1M tokens
Input	$ 0.33	$ 0.33>= 512K$ 0.66
Output	$ 1.32	$ 1.32>= 512K$ 2.63
Cache read	$ 0.066	$ 0.066>= 512K$ 0.13
Cache write	$ 0.42	$ 0.33>= 512K$ 0.66
What it is good at
Best for	Complex agent workflows Repository-scale coding	Multimodal agents Computer use and cowork Long-context research
Published benchmarks
Terminal-Bench 2.0pass_rate	57%(setup not fully disclosed)	—
Terminal-Bench 2.1pass_rate	—	66%Terminus 2(setup not fully disclosed)
Multi-SWE-Benchresolved_rate	52.7%(setup not fully disclosed)	—
SWE-Bench Proresolved_ratesetups differ by: harness	56.22%(setup not fully disclosed)	59%Claude Code(setup not fully disclosed)
SWE-Bench Verifiedresolved_rate	—	80.5%(setup not fully disclosed)
GDPval-AAelo	1495(setup not fully disclosed)	—
MMMU Pro (no tools)accuracy	—	78.1%(setup not fully disclosed)
MCP Atlas Publicscore	—	74.2%(setup not fully disclosed)
Toolathlonscore	46.3%(setup not fully disclosed)	—