MiniMax M3
minimax/minimax-m3MiniMax-M3 is MiniMax's flagship multimodal foundation model, which MiniMax describes as open-weight. It is built on MiniMax Sparse Attention, which selects key-value blocks to cut long-context compute and supports up to a 1M-token context, and it was trained natively on text and visual data for image and video understanding. MiniMax positions it for long-horizon coding, agentic tool use and autonomous multi-step work such as research paper reproduction and GPU kernel optimization.
- Input price
- $0.22
$0.30USD / 1M tokens - Output price
- $0.90
$1.20USD / 1M tokens - Context window
- 1M
- Max output
- 512K
Try the model
Playground
Try a prompt
Your response will appear here
Choose an example or write a prompt, then click Run.
Model card
Specifications
Pricing
- Input
- $0.22 / 1M tokens
- Output
- $0.90 / 1M tokens
- Cache read
- $0.04 / 1M tokens
Context & modalities
- Context window
- 1,000,000 tokens
- Max output
- 512,000 tokens
- Input
- Text, image
- Output
- Text
Capabilities
- Chat
- Supported
- Vision
- Supported
- Reasoning
- Supported
- Structured output
- Supported
- Function calling
- Supported
- Audio input
- Not supported
Access
- Provider
- MiniMax
- Model ID
- minimax/minimax-m3
- Execution
- sync
- Base URL
- https://api.sandbase.ai
- API
- Chat Completions API
- Endpoint
- /v1/chat/completions
Independent evaluations
Benchmarks
Scores on standardized evaluations, compared with other models on SandBase.
Metrics sourced from Artificial Analysis · Intelligence Index v5
Start building
Send your first request
OpenAI-compatible endpoint with unified authentication and usage tracking.
# The quoted heredoc keeps Unicode and shell metacharacters unchanged.
result=$(curl --fail-with-body --silent \
-X POST "https://api.sandbase.ai/v1/chat/completions" \
-H "Authorization: Bearer $SANDBASE_API_KEY" \
-H "Content-Type: application/json" \
--data-binary @- <<'SANDBASE_JSON'
{
"model": "minimax/minimax-m3",
"messages": [
{
"role": "user",
"content": "Hello"
}
],
"max_tokens": 2048
}
SANDBASE_JSON
)
printf '%s\n' "$result"Choose your model
Compare models
| Model | Context | Input / 1M | Output / 1M | Released |
|---|---|---|---|---|
MiniMax M3This model MiniMax | 1M | $0.22 | $0.90 | May 31, 2026 |
MiniMax | 1M | $0.15 | $0.60 | Oct 19, 2026 |
MiniMax | 200K | $0.30 | $1.20 | Mar 18, 2026 |
MiniMax | 200K | $0.30 | $1.20 | Feb 12, 2026 |
Questions
FAQ
How do I call MiniMax M3 through SandBase?
Create a SandBase API key, then send requests with the model ID "minimax/minimax-m3" to Chat Completions API (/v1/chat/completions) at https://api.sandbase.ai. The request examples on this page show the exact payload.
How much does MiniMax M3 cost?
$0.22 per 1M input tokens and $0.90 per 1M output tokens with the current 25% discount (list price $0.30 / $1.20), billed per request from your SandBase balance.
What is the context window of MiniMax M3?
1M tokens of context, with up to 512K output tokens per response.
How does MiniMax M3 score on benchmarks?
MiniMax M3 scores 18.7 on the Artificial Analysis Intelligence Index (v5), ranking #20 of 21 models on SandBase. Source: Artificial Analysis, measured as MiniMax-M3.
Which features does MiniMax M3 support?
MiniMax M3 supports vision (image input), reasoning, function calling, structured output. See Specifications above for the full list.
Do I need a separate MiniMax account?
No. One SandBase API key and balance gives you access to MiniMax M3 and the other models in the catalog; you do not need to sign up with MiniMax separately.