MiniMax M3
MiniMax · Open weights · Released May 31, 2026 · minimax/minimax-m3
1.05M contextVisionToolsReasoningStructured outputPrompt cachingOpen weights
Available from 13 providersUSD / MTOK · VERIFIED 12 HOURS AGO
EST. MONTHLY COSTclick a row for detail · click headers to sort
| PROVIDER | INPUT | OUTPUT | CACHE | CONTEXT | QUANT | TTFT | TPS | SMOKE | UPTIME |
|---|---|---|---|---|---|---|---|---|---|
| CWCoreWeave | $0.23 | $0.96 | $0.05 | 262K | fp4 | 0.93s | 98 tok/s | 4/5 | 99.9% |
| GMIGMICloud | $0.24 | $0.96 | $0.048 | 1.05M | fp8 | 2.5s | 136 tok/s | 5/5 | 99.8% |
| MOMorph | $0.255 | $1.02 | — | 256K | fp4 | 2.2s | 50 tok/s | 3/5 | 93.9% |
 DeepInfra | $0.28 | $1.10 | $0.056 | 524K | fp8 | 0.48s | 19 tok/s | 4/5 | 98.6% |
| ATAtlasCloud | $0.30 | $1.20 | $0.06 | 524K | fp8 | 0.81s | 57 tok/s | 4/5 | 99.6% |
 Parasail | $0.30 | $1.20 | $0.06 | 1.05M | fp8 | 1.5s | 55 tok/s | 3/5 | 98.6% |
 Novita | $0.30 | $1.20 | $0.06 | 1M | fp8 | 1s | 127 tok/s | — | 99.7% |
| STStreamLake | $0.30 | $1.20 | $0.06 | 1M | fp8 | 1.7s | 195 tok/s | — | 99.6% |
 Together | $0.30 | $1.20 | $0.06 | 524K | — | 0.37s | 188 tok/s | — | 99.2% |
 Venice | $0.30 | $1.20 | $0.06 | 524K | fp8 | 0.82s | 173 tok/s | — | 99.9% |
 Minimax1ST PARTY | $0.30 | $1.20 | $0.06 | 524K | fp8 | 1.1s | 114 tok/s | — | 99.4% |
 SambaNova | $0.60 | $2.40 | — | 1.05M | — | 1.8s | 263 tok/s | — | 98.9% |
| MOModelRun | $0.75 | $3.00 | $0.15 | 1.05M | fp4 | 0.39s | 230 tok/s | — | 99.7% |
TTFT/TPS: median (P50) over trailing 72h · chat10k profile (~10K in / 200 out) · measured via OpenRouter routing · SMOKE: of 5 programmatic capability probes passed (JSON schema, instruction following, tool call, long-context, benign compliance; retry-once) · screened via OpenRouter routing · v0.1 · methodology v1.1 · measured 43 hours ago
Pricing variants · USD / MTok in / out
STANDARD
$0.23 / $0.96
per MTok in / out
BATCH
$0.30 / $1.20
async batch
CACHE READ $0.05
About
MiniMax-M3 is a multimodal foundation model from MiniMax. It supports text, image, and video inputs with text output, a 1M-token context window, and is suited for long-horizon agentic work, coding,...
Specifications
Context window1,048,576
Max output512,000
Modalities intext, image, video
Modalities outtext
LicenseOpen weights
ReleasedMay 31, 2026
TokenizerOther
WeightsHugging Face ↗
Related
Embed badge
[](https://modelindex.ai/models/minimax/minimax-m3)Sources
OpenRouter API12 hours ago
AGGREGATORHugging Face12 hours ago
COMMUNITYModelIndex benchmarks43 hours ago
MEASUREDVerified 12 hours ago · list prices, not negotiated rates.