LANGUAGE & CODING / MiniMax
MiniMax M3
Multimodal coding and agent tasks with a million-token context.
PRICE / USD
$0.3
Per 1 million output tokens · input billed separately
- Input / 1M tokens
- $0.3
- Output / 1M tokens
- $1.2
- Cached input / 1M tokens
- $0.06
- Cache write / 1M tokens
- Not verified
Above 512K input, token rates double. Priority processing costs 1.5× the Standard rate.
MiniMax pay-as-you-go pricing ↗SPEED & CAPABILITIES
Not benchmarked
Provider description: ≈100+ tokens/s. This does not qualify for the numeric (fast) badge.
MiniMax's approximate output-speed claim; workload, tokenizer, and latency are not matched to other providers.
MiniMax language models ↗- Model score
- 29
- Our own output-speed measurement
- Not measured
- Context window
- 1,000,000
- API model ID
MiniMax-M3
Artificial Analysis tested MiniMax M3 · reasoning. This mostly English, text-based index is not a guarantee of performance on your workload. Artificial Analysis · MiniMax M3 · reasoning ↗
M3.1 Flash Preview is listed separately by MiniMax as subscription-only; these prices apply to M3 pay-as-you-go.
Suggested starting role: everyday coding. Low listed rates plus documented coding, agent, and visual capabilities make M3 a candidate for everyday development. Validate tool calls and code changes against your own checks. This is editorial guidance, not a benchmark score.
Sources & methodology
MiniMax language models ↗ Checked 2026-10-05
MiniMax pay-as-you-go pricing ↗ Checked 2026-10-05
Artificial Analysis · MiniMax M3 · reasoning ↗ Checked 2026-10-05
Artificial Analysis · Intelligence Index methodology v4.3.2 ↗ Checked 2026-10-05