MiniMax Open Platform
MiniMax M2.1 pricing
Leverage the faster M2.1 variant for pipeline orchestration while keeping $0.30/$1.20 per-million spend predictable.
Last pricing check: Mar 13, 2026
Why teams choose this model
Scenario planning
Realistic cost examples
Numbers use MiniMax M2.1 pricing
Multi-agent coding lane
Cursor-style agents plan, code, run tests, and patch PRs with 200K context.
Per request
$0.0024
Per month
$19.20
Tokens sent
28,000,000
Meeting analyst
M2.1 ingests transcripts, hits internal tools, and outputs action plans.
Per request
$0.0017
Per month
$20.88
Tokens sent
33,600,000
Agent orchestration layer
Ops teams run tool-rich playbooks with the high-speed SKU for lower latency.
Per request
$0.0015
Per month
$30.60
Tokens sent
48,000,000
Compare with
FAQs
How is M2.1 different from M2?
Same token pricing + context, but MiniMax advertises ~100 tps output speeds and stronger coding/agent benchmarks, so we tag it as a premium latency class.
Can I mix Bedrock and MiniMax’s native API?
Yes—the Bedrock pricing we use matches MiniMax’s own $0.30/$1.20 rates, so switching providers shouldn’t change totals.
Any batch discounts?
Bedrock Batch cuts the rate in half, but we default to on-demand numbers in TokenTally until we add multi-tier toggles.
Pricing sources
- https://aws.amazon.com/bedrock/pricing/
Checked Mar 13, 2026
- https://platform.minimax.io/docs/api-reference/text-intro
Checked Mar 13, 2026