Official site ◆ · 🇨🇳 China
| Tier | $/month | Included usage |
|---|---|---|
| Plus | $20.00 | — |
| Max | $50.00 | — |
| Ultra | $120.00 | — |
readable by API units unstated
As advertised. no numeric quota published; console shows a usage bar
What that actually means. No token, credit, prompt or dollar figure for any tier -- only '3-4 agents'. Verified directly: the API returns a remaining percentage for text with total_count 0, so even the endpoint hides the denominator. Video quota does return absolute counts.
Where to check what is left. GET /v1/token_plan/remains, plus the console bar
Reverse-engineered. This vendor publishes no figure, so one was derived: about 7.4B tokens a week on Ultra ($120/month), roughly 32B a month, or $0.0038 per million. Derived from a SINGLE account; not an official figure. Robust to the API's integer-percentage resolution (worth about 0.05B) and to the rolling-versus-fixed window mismatch (7.2-7.7B). One caveat that cannot be resolved from outside: the internal weighting of cached input, fresh input and output has not been reverse-engineered, so this figure is valid at the observed ~92% cache-hit rate and should not be extrapolated to a very different one. Note also that the accounting has already changed once: cached tokens formerly did not count against the quota and now do, which raised the effective cost sharply and was not announced.
checked 2026-08-13 · high (price) / low (quota) · source
| Model | Coding index | tok/s | $/M in | $/M out | $ / session | Cache |
|---|---|---|---|---|---|---|
| MiniMax-M3 | 58.6 | 70.0 | $0.30 | $1.20 | $0.5434 | 94.3% |
| MiniMax M2.7 | 52.6 | 0.0 | $0.30 | $1.20 | — | — |
3 endpoints, cheapest first. This is the view the model-by-model tables cannot give you: what one provider actually offers, and at what precision.
| Model | Quant | $/M in | $/M out | Context | Uptime 30m |
|---|---|---|---|---|---|
| MiniMax-M3 | fp8 | $0.30 | $1.20 | 524,288 | 97.6% |
| MiniMax M2.7 | fp8 | $0.30 | $1.20 | 204,800 | 97.6% |
| MiniMax M2.7 | fp8 | $0.60 | $2.40 | 204,800 | — |
Everyone serving these models, cheapest first, with the numeric precision each one runs. The lab is often not the cheapest place to buy its own model.
| Model | Served by | Quant | $/M in | $/M out | Uptime 30m |
|---|---|---|---|---|---|
| MiniMax-M3 | CoreWeave | fp4 | $0.23 | $0.96 | 99.9% |
| MiniMax-M3 | GMICloud | fp8 | $0.24 | $0.96 | 97.8% |
| MiniMax-M3 | DeepInfra | fp8 | $0.28 | $1.10 | 98.8% |
| MiniMax-M3 | AtlasCloud | fp8 | $0.30 | $1.20 | 99.7% |
| MiniMax-M3 | MiniMax | fp8 | $0.30 | $1.20 | 97.6% |
| MiniMax-M3 | Morph | fp4 | $0.30 | $1.20 | 96.7% |
| MiniMax-M3 | Novita | fp8 | $0.30 | $1.20 | 99.7% |
| MiniMax-M3 | Parasail | fp8 | $0.30 | $1.20 | 99.0% |
| MiniMax-M3 | StreamLake | fp8 | $0.30 | $1.20 | 99.5% |
| MiniMax-M3 | Together | unknown | $0.30 | $1.20 | 99.5% |
| MiniMax-M3 | Venice | fp8 | $0.30 | $1.20 | 99.8% |
| MiniMax-M3 | ModelRun | fp4 | $0.75 | $3.00 | 99.9% |
| MiniMax M2.7 | Mara | unknown | $0.24 | $0.96 | 100.0% |
| MiniMax M2.7 | DeepInfra | fp8 | $0.25 | $1.00 | 100.0% |
| MiniMax M2.7 | GMICloud | fp8 | $0.27 | $1.08 | 100.0% |
| MiniMax M2.7 | Novita | fp8 | $0.27 | $1.08 | 98.2% |
| MiniMax M2.7 | AtlasCloud | fp8 | $0.30 | $1.20 | — |
| MiniMax M2.7 | Fireworks | unknown | $0.30 | $1.20 | — |
| MiniMax M2.7 | MiniMax | fp8 | $0.30 | $1.20 | 97.6% |
| MiniMax M2.7 | DeepInfra | fp8 | $0.38 | $1.70 | 99.5% |
| MiniMax M2.7 | Groq | unknown | $0.60 | $1.80 | 100.0% |
| MiniMax M2.7 | MiniMax | fp8 | $0.60 | $2.40 | — |
| MiniMax M2.7 | SambaNova | unknown | $0.60 | $2.40 | — |