ModelPriceWatch.com
Last scan 2026-08-01 Models tracked 181 Providers 27 Cheapest paid Granite 4.0 Micro $0.017/Mtok in Every price links to its source

MiniMax-M2.5

by MiniMax

Retired budget open weights cheap tier
Email me when MiniMax-M2.5 gets cheaper
Retired on Jul 28, 2026. No longer served on the endpoint we price — the figures below are kept for historical reference only. Replaced by MiniMax-M2.7.
Availability: MiniMax's Pay-as-You-Go pricing page (checked 2026-07-28) lists MiniMax-M2.5 under 'Legacy Models', not among its actively-offered text LLMs (which are MiniMax-M3 and MiniMax-M2.7). The $0.30/$1.20 shown is a real historical price — it still appears on the Legacy row and LiteLLM carries minimax/MiniMax-M2.5 at the same value — but M2.5 is no longer the current offering, so treat it as superseded by MiniMax-M2.7. This is a lifecycle reclassification; the price itself did not change.

Today's price · per 1M tokens

Input

$0.300

per 1M tokens

Output

$1.20

per 1M tokens

Blended

$0.750

avg of input & output $/1M

Cached input

$0.030

10% of input — prompt caching

Source: official MiniMax pricing · read Jul 28, 2026 MODELPRICEWATCH.COM · 2026-08-01
MiniMax M2.5 is still sold by Fireworks at $0.750 blended per 1M tokens (host details →)

Overview

MiniMax M2.5 model. $0.30/$1.20 per 1M; cached $0.03. Superseded by MiniMax-M2.7.

Run it yourself

Deploy this open model on rented GPUs

Open weights mean you can self-host instead of paying per-token API prices. These platforms let you serve it on demand — often cheaper at scale.

RunPod →
Rent GPUs by the second and serve any open-weight model yourself.
Novita AI →
Serverless open-model inference billed per token — no GPU ops.
Vast.ai →
GPU marketplace — rent spot/on-demand GPUs cheaply and serve any open-weight model.

Partner links — we may earn a commission if you sign up, at no cost to you. We only list platforms we'd recommend regardless.

Capabilities

struck through = not supported
Input 1/5
Text ✓ Image Audio Video PDF
Output 1/5
Text ✓ Image Audio Video Embedding
Features 3/9
Prompt caching ✓ Reasoning Coding Fast inference Long context ✓ Open weights ✓ Multimodal Web search Realtime

Benchmark performance

accuracy % · higher is better
Percentile vs all tracked models 23.3th2 independent measurements
Percentile per $/Mtok 31.1

We hold no per-benchmark accuracy scores for this model yet, so it has no accuracy average. That is a gap in our coverage, not a sign the model is untested — independent evaluators often publish a composite index for a new model long before releasing its per-benchmark numbers. The percentile above is its standing across the independent composites below.

Independent composite scores
  • AA-Omniscience Index: -39.7 (−100–100)
  • LMSYS Chatbot Arena: 1389.9 ELO (Human preference)
How it stacks up
  • Ranks #98 of 133 comparably-measured models by percentile score, across 2 independent measurements
  • Ranks #49 of 133 comparably-tested models by normalized performance per dollar

Source: LMSYS Chatbot Arena (UC Berkeley) · updated Jun 29, 2026 · See full rankings →

Specifications

Provider
MiniMax
Context window
205K tokens
Modality
text
Parameters
Proprietary
Open source
Yes — open weights available
Released
Jul 1, 2025
Status
Retired ended Jul 28, 2026
Last updated
Jul 28, 2026
Tags
open-weightsbudgetcachingretired

Availability verified: Jul 28, 2026 — per MiniMax's own deprecation notice

Try on MiniMax Official pricing page