MiniMax: MiniMax M3
by MiniMaxMiniMax: MiniMax M3 is a large language model from MiniMax. It costs $0.300 per million input tokens and $1.20 per million output tokens. Its context window is 1.0M tokens.
- Input / 1M tokens
- $0.300
- Output / 1M tokens
- $1.20
- Cached input / 1M
- $0.060
- Context window
- 1.0M
On repeated prefixes
tokens
Who serves it cheapest
12 hosts serve MiniMax M3. Same weights, same API - the price difference is pure margin and routing.
| Provider | Input / 1M | Output / 1M | Context | Throughput | Uptime 24h |
|---|---|---|---|---|---|
| CoreWeaveCheapestfp4 | $0.230 | $0.960 | 262K | - | 100.0% |
| GMICloudfp8 | $0.240 | $0.960 | 1.0M | - | 99.5% |
| DeepInfrafp8 | $0.280 | $1.10 | 524K | - | 96.4% |
| StreamLakefp8 | $0.300 | $1.20 | 1M | - | 98.6% |
| Venicefp8 | $0.300 | $1.20 | 524K | - | 95.5% |
| Together | $0.300 | $1.20 | 524K | - | 99.0% |
| Parasailfp8 | $0.300 | $1.20 | 1.0M | - | 99.2% |
| AtlasCloudfp8 | $0.300 | $1.20 | 524K | - | 98.8% |
| Novitafp8 | $0.300 | $1.20 | 1M | - | 99.6% |
| Minimaxfp8 | $0.300 | $1.20 | 524K | - | 98.5% |
| SambaNova | $0.600 | $2.40 | 1.0M | - | 99.4% |
| ModelRunfp4 | $0.750 | $3.00 | 1.0M | - | 99.9% |
The spread between CoreWeave and ModelRun is 3.2× for identical weights. Quantization and context limits differ, so check both columns before switching.
Benchmarks
Independent scores published alongside the catalogue.
- Intelligence index
- 29.6
- Coding index
- 58.6
- Agentic index
- 30.8
About MiniMax M3
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
| Model ID | minimax/minimax-m3 |
|---|---|
| Provider | MiniMax |
| Context window | 1.0M tokens |
| Max output | 512K tokens |
| Input modalities | text, image, video |
| Output modalities | text |
| Knowledge cutoff | - |
| Open weights | Yes - MiniMaxAI/Minimax-M3 |
| Released | May 31, 2026 |
Cheaper alternatives
Models that cost less than MiniMax M3 while keeping at least half its context window and every input modality it supports.
Frequently asked
How much does MiniMax: MiniMax M3 cost?
$0.300 per million input tokens and $1.20 per million output tokens. Cached input reads cost $0.060 per million tokens.
What is the context window of MiniMax: MiniMax M3?
1.0M tokens, with up to 512K tokens of output per request.
Which provider serves MiniMax: MiniMax M3 cheapest?
CoreWeave at $0.230 per million input tokens - 3.2× cheaper than ModelRun, the most expensive of the 12 hosts serving it.
Confirm against the source: MiniMax official pricing.