LLM Pricing
PricingLeaderboardsToolsProvidersReleasesGuides

© 2026 LLM Pricing

About
·Contact
·Privacy
·RSS
← Model list

MiniMax-M2.5

minimax·minimax/MiniMax-M2.5·Deprecated·Open weights·minimax series
Prior MiniMax coding model for agent workflows, office edits, and automation

MiniMax-M2.5 is offered by 42 providers on this page. Public prices are shown for 33 of them. Its official list price is $0.30 per 1M input tokens and $1.20 per 1M output tokens. The lowest paid channel is DInference at $0.22 / $0.88 per 1M, about 2.0× below the list price. That channel is a third-party gateway, so confirm its availability and rate limits before depending on it. It also has 8 covered by a paid subscription; free tiers usually carry rate limits, and subscription-covered access bills $0 per token only after the subscription fee.

Artificial Analysis rates it 34.5 on the Intelligence Index. Against its lowest blended price of $0.385 per 1M, that is roughly 90 index points per dollar, which is the value ratio the leaderboards rank on. Median output speed is 90 tokens per second, with 1.68s to the first token.

The context window is 204,800 tokens at the reference host, but hosts report different limits, from 192,000 to 228,700, so the usable window depends on the provider you pick. It supports reasoning, tool use, and structured output. The weights are open, so it can also be self-hosted or served through a gateway of your choice. Providers report the capability flags inconsistently, so verify a specific feature against the host you plan to use. Its training knowledge cuts off at 2025-01.

Specs & pricing

Input / output per 1M tokens
Official price·MiniMax (minimax.io)
$0.30 / $1.20
Blended $0.525 · Cache read $0.03
Lowest paid·DInferenceGateway
$0.22 / $0.88
Blended $0.385 · 2× spread
8 more $0 channels
Context
204,800
Output limit
131,072
Knowledge cutoff
2025-01
Released / updated
2026-02-12 / 2026-02-12
Capabilities
✓ Reasoning✓ Tool use✓ Structured output✓ TemperatureAttachments
⚠ Providers report capability flags inconsistently
Modalities
Text

Quality & performanceArtificial Analysis · Intelligence Index v4.1

Intelligence34.5
Value90
Coding—
Agentic—
Output speed90 tok/s
TTFT1.68 s
Value formulaIQ 34.5 ÷ min blended $0.385 = 90

Quality is independently evaluated by Artificial Analysis. Speed/latency are model-level medians.

Available at 42 providers33 with public prices · 8 subscription-covered

ProviderTierInputOutputCache readCache writeContextOutput limitStatus
Alibaba Token PlanGatewaySubscription$0$0196,608 ⚠32,768
Alibaba Coding Plan (China)GatewaySubscription$0$0196,608 ⚠24,576
MiniMax Token Plan (minimaxi.com)GatewaySubscription$0$0204,800131,072
MiniMax Token Plan (minimax.io)GatewaySubscription$0$0204,800131,072
Alibaba Coding PlanGatewaySubscription$0$0196,608 ⚠24,576
SCNet Token PlanGatewaySubscription$0—204,800131,072
Tencent Coding Plan (China)
minimax-m2.5
GatewaySubscription$0$0204,80032,768
Alibaba Token Plan (China)GatewaySubscription$0$0196,608 ⚠32,768
DInference
minimax-m2.5
Gateway$0.22$0.88——200,000 ⚠32,000
Deep InfraCloud$0.15$1.15$0.03—196,608 ⚠131,072deprecated
MiniMax (minimax.io)OfficialFirst-party$0.30$1.20$0.03$0.375204,800131,072

Sorted by blended price (input×0.75 + output×0.25) asc. The official channel always shows regardless of rank. Whether a gateway's low price is actually usable can't be verified.

Your usage cost

1Venice AI$72.70
2Deep Infra$73.10
3OpenRouter$78.84
4GreenPT$79.48
5TokenGo$87.60
6OrcaRouter$87.60
9MiniMax (minimax.io) · Official$87.60

8 more channels offer $0 (Alibaba Token Plan, Alibaba Coding Plan (China), MiniMax Token Plan (minimaxi.com) etc.); free tiers usually have rate limits and no SLA, excluded from ranking.

Switch to Venice AI to save $14.90/mo (17%). The gap is small, so staying on the official channel is fine.
Note: this is a gateway; verify availability and rate limits yourself.

Benchmark4 items

NameConditionsScoreMetricSource
SWE-Bench Verified—75.8resolvedSource ↗
SWE-Atlas Codebase QnAharness: Mini-SWE-Agent10.3scoreSource ↗
SWE-Atlas Refactoringharness: Mini-SWE-Agent19.52scoreSource ↗
SWE-Atlas Test Writingharness: Mini-SWE-Agent18.6scoreSource ↗

The same benchmark scores very differently across harness / dataset, so the qualifying conditions must be shown together.

Reasoning control

effort = noneeffort = minimaleffort = loweffort = mediumeffort = higheffort = xhigheffort = max

36 / 42 providers expose no reasoning control (reasoning_options: []).

Related models

MiniMax H3 Maxsame series—MiniMax H3same series—MiniMaxAI/MiniMax-M2.5same series$0.30 / $1.22DeepSeek V4 Flashcheaper alternative$0.14 / $0.28GLM-5.3-Flashcheaper alternative$0.075 / $0.25GPT-5 Nanocheaper alternative$0.05 / $0.40MiMo-V2.5cheaper alternative$0.14 / $0.28

Price historyone sample accumulated per data sync

Input listOutput listMin blended
$1.20$02026-08-052026-08-232026-08-05 · Input list $0.302026-08-06 · Input list $0.302026-08-07 · Input list $0.302026-08-08 · Input list $0.302026-08-09 · Input list $0.302026-08-10 · Input list $0.302026-08-11 · Input list $0.302026-08-12 · Input list $0.302026-08-13 · Input list $0.302026-08-23 · Input list $0.302026-08-05 · Output list $1.202026-08-06 · Output list $1.202026-08-07 · Output list $1.202026-08-08 · Output list $1.202026-08-09 · Output list $1.202026-08-10 · Output list $1.202026-08-11 · Output list $1.202026-08-12 · Output list $1.202026-08-13 · Output list $1.202026-08-23 · Output list $1.202026-08-05 · Min blended $0.322026-08-06 · Min blended $0.322026-08-07 · Min blended $0.322026-08-08 · Min blended $0.322026-08-09 · Min blended $0.322026-08-10 · Min blended $0.322026-08-11 · Min blended $0.322026-08-12 · Min blended $0.322026-08-13 · Min blended $0.322026-08-23 · Min blended $0.385
Data partly from models.dev (MIT) · MiniMax (minimax.io) official docs ↗