Calculator
MiniMax M2.5: self-host or API?
Self-hosting
No throughput figureThe API is cheaper below 545 million tokens a day. Above that, self-hosting wins only if the whole configuration (all GPUs together) delivers at least 12,627 tokens per second, input and output together, while serving at 50% utilization (6,314 if it were busy all the time). For comparison, published figures decide it for a dearer setup: 4x B200 SXM 180GB at $652 a day, self-hosting cheaper above 1,481 million tokens a day (published minimum, for requests of up to 2,048 tokens in total).
At 50% utilization · Cheapest API offer: $0.44 per million tokens at Venice (3:1 input:output mix)
Hardware rental only: the hourly price the provider publishes for the machine. It does not count the engineering to deploy and run the model, monitoring, redundancy, start-up time, or storage and network costs billed separately.
- Break-even volume
- 545.45 million tokens a day
- Throughput self-hosting needs
- 12,627 tokens per second
- At 50% utilization
- Throughput self-hosting needs
- 6,314 tokens per second
- If busy all the time
At least this many input plus output tokens per second, all GPUs together, while serving.
What published figures show
No throughput figure for this model on this configuration.
Rented configuration
8 x L40 48GB at CoreWeave
$240.00 a day · $10.00 an hour · Guaranteed · 384 GB of GPU memory · fp8
Checked Oct 31, 2026 Price page at the provider (opens in a new tab)
For comparison: published figures decide it for a dearer setup
4 x B200 SXM 180GB at Lambda
$651.84 a day · $27.16 an hour · Guaranteed
Self-hosting is cheaper above the break-even volume · 1,481.45 million tokens a day
Published minimum Checked Oct 31, 2026 Price page at the provider (opens in a new tab)
Another, dearer setup, shown for comparison. It is not a recommendation.
Calculation detail
Cheapest of 24 current standard offers, promotions, unstable prices, off-peak prices without published hours and unverified prices far below those of the other providers excluded Checked Oct 31, 2026
Assumptions
- We have no measured, derived or estimated throughput for this model on this card, and we have not assessed inference-engine support for it.
- Hardware rental only: the hourly price the provider publishes for the machine. It does not count the engineering to deploy and run the model, monitoring, redundancy, start-up time, or storage and network costs billed separately.
- Configuration: we pick the cheapest one whose weights fit. If a configuration costing up to 25% more has a measured throughput, we pick it instead; failing that, one with a derived throughput; failing that, one with a published lower bound that settles the verdict; among equals, the cheaper. A configuration whose throughput is only a wide estimate is never preferred over the cheapest. This is our rule, not a technical requirement.
- Quantization fp8; guaranteed tier prices as published by the provider; utilization 50%.
- Throughput is aggregate (input and output tokens, 3:1). Published figures in generated tokens are converted with a prefill/generation speed ratio of 3 (prudent), 5 (median) and 8 (optimistic).
- Break-even = daily GPU cost / API price per million tokens (545 million tokens a day). Self-hosting is cheaper only if the average utilization you can sustain exceeds break-even / raw capacity (the break-even utilization of each scenario), whatever the number of GPUs.