← Back to model catalog
MiniMax

MiniMax M2.5

minimax-m2.5
Guru+coding

MiniMax M2.5 is a coding model in the Majestix lineup built for coding, agentic, tool use workloads, with 128K context and $0.32 / $1.00 per 1M tokens.

Context Window
128K
Max Output
32K
Pricing
$0.32 input / $1.00 output per 1M tokens

What this model is good at

CodingAgenticTool UseReasoning

Benchmarks and operating signals

SignalValueNote
Context Window128Kplatform limit
Max Output32Kplatform limit
Output Price$1.00 / 1MMajestix pricing

Quick guidance

Use MiniMax M2.5 when your workload lines up with coding and agentic and you want a clear price-to-capability tradeoff inside Majestix.

This card intentionally focuses on platform context: pricing, limits, and where the model already appears in swarm templates. It is not a self-hosting guide or a vendor deployment manual.

If you want the narrative version, the paired blog intro is the fastest way to explain this model to customers, internal operators, or community readers.