MiniMax M3: API Pricing, Context Window & Benchmarks
MiniMax M3 is a language model from MiniMax, released in June 2026, with multimodal input, a 1M-token context window, and pricing from $0.300/M input, $0.059/M cached input, $1.20/M output.
MiniMax M3 is the first open-weight model to combine three frontier capabilities: top-tier coding and agentic performance, a 1M-token context window, and native multimodality. It is powered by MiniMax Sparse Attention (MSA), a new sparse
MiniMax M3 benchmarks
Capability tiers
Standing within each category, adjusted for leaderboard depth.
Real tasks performance
High-confidence performance for MiniMax M3 across real-world prompt categories. Only 95% intervals at most 4 points wide are shown.
Performance by conversation depth
How MiniMax M3 holds up as conversations get longer.
Quality Tracker
MiniMax M3 Performance Across Datasets
Scores sourced from the model's scorecard, paper, or official blog posts
llm-stats.com - Sat Sep 05 2026MiniMax M3 pricing
Providers
MiniMax M3 starts at $0.300 per million input tokens and $1.20 per million output tokens via Fireworks. Reused prompt prefixes cost $0.0590 per million cached input tokens. See all 4 providers below with their per-token pricing, latency, throughput, and modality support.
| Provider | Input $/M | Cached input $/M | Output $/M | Context in / out | TTFT p95 s | Output p5 c/s | Modalities in / out |
|---|---|---|---|---|---|---|---|
 Fireworks | $0.300 | $0.0590 | $1.20 | 512.0K/131.1K | — | — | / |
 MiniMax | $0.300 | $0.0600 | $1.20 | 1.0M/524.3K | 3.03 | 13 | / |
 Novita | $0.300 | $0.0600 | $1.20 | 1.0M/131.1K | — | — | / |
 Together | $0.300 | $0.0600 | $1.20 | 1.0M/131.1K | — | — | / |
Cached input is the discounted price for prompt tokens served from a provider cache. TTFT is time to first token. Output is characters per second; p5 is the sustained floor exceeded by 95% of observed requests.
MiniMax M3 model size
MiniMax M3 has 428 billion parameters. See how it compares to other models in the same parameter range.
MiniMax M3 context window
Input and output token limits for MiniMax M3, plus how it ranks on long-context understanding.
MiniMax M3 API
Run a request to see the response
Use it in your code
Billed at $0.30 input / $0.06 cached input / $1.20 output per 1M tokens through the LLM Stats gateway.
from openai import OpenAI
client = OpenAI(
api_key="YOUR_API_KEY",
base_url="https://gateway.llm-stats.com/v1"
)
response = client.chat.completions.create(
model="minimax-m3",
messages=[
{"role": "user", "content": "What is machine learning?"}
]
)
print(response.choices[0].message.content)Need an API key? Create one above in the playground, or read the API documentation.
MiniMax M3 latency
MiniMax M3 time to first token, sustained output throughput, and failed-request rate from live API traffic over the trailing 7 days.
Provider operational metrics
Time to first token, output throughput, and failed-request rate from live API traffic
MiniMax M3 examples
Recent arena outputs from MiniMax M3, picked from the highest-ranked matchups.
MiniMax M3 license
MiniMax M3 is released under the MIT license, which permits commercial use, has 428.0B parameters.
- License
- MIT
- Commercial use allowed
- Parameters
- 428.0B
MIT License - allows commercial use
MiniMax M3 resources
Official sources for MiniMax M3: api documentation, official playground, official launch post, model weights.
MiniMax M3 vs other models
The most-compared alternatives to MiniMax M3 are Claude Opus 4.6, Qwen3.7 Max, GPT-5. Open any pair side-by-side for benchmarks, pricing, context, and latency.
Models like MiniMax M3
Models ranked just above and below MiniMax M3 by LLM Stats score.
FAQ
Common questions about MiniMax M3.