Models / MiniMax / MiniMax-M2.5

MiniMax · minimax

MiniMax-M2.5

Prior MiniMax coding model for agent workflows, office edits, and automation

$0.30 in · $1.20 out

per 1M tokens · first-party price. Reference price from MiniMax, the lab's own listing.

Provider input prices run from $0.19 to $0.35 per 1M tokens. Output prices run from $0.88 to $1.39.

Context
204,800
Max output
131,072
Cache read
$0.03
Cache write
$0.38
Reasoning
Yes
Tool calls
Yes
Structured output
Unknown
Vision
No
Open weights
Yes
License
—
Released
Feb 12, 2026
Updated in source
Feb 12, 2026
Knowledge
—
Providers
25

Hugging Face weights

What this model costs per month

Prices are USD per 1M tokens. A month is 30 days. Cache hits use the cache-read rate when the model has one; otherwise the cache field is ignored.

monthly = requests/day × 30 × (input tokens × (1 − cache hit) × input price + input tokens × cache hit × cache-read price + output tokens × output price) ÷ 1,000,000

Providers

Base rates from Models.dev. Where a provider lists the model more than once, the row is the listing whose id is closest to the canonical model id.
ProviderInputOutputCache readCache writeStatus
GreenPT$0.19$1.13$0.063——
DInference$0.22$0.88———
OpenRouter$0.27$1.08$0.027——
Venice AI$0.27$0.95$0.03——
302.AI$0.30$1.20———
Amazon Bedrock$0.30$1.20———
DevPass$0.30$1.20$0.03$0.38—
Eden AI$0.30$1.20$0.03——
Friendli$0.30$1.20$0.06——
HPC-AI$0.30$1.20$0.03——
Kilo Gateway$0.30$1.20$0.03——
LLM Gateway$0.30$1.20$0.03——
Merge Gateway$0.30$1.20$0.03$0.38—
MiniMaxreference$0.30$1.20$0.03$0.38—
NanoGPT$0.30$1.20$0.15——
Novita AI$0.30$1.20$0.03——
Ofox$0.30$1.20$0.03$0.38—
OrcaRouter$0.30$1.20$0.03$0.38—
TensorX$0.30$1.20$0.075$0.38—
TokenGo$0.30$1.20$0.03——
Vercel AI Gateway$0.30$1.20$0.03$0.38—
Cortecs$0.35$1.39$0.075——
Alibaba Token Plan$0$0$0$0—
Alibaba Token Plan$0$0$0$0—
SCNet Token Plan$0$0$0——