MiniMax M2.5 vs MiniMax M3
MiniMax M2.5 (MiniMax) versus MiniMax M3 (MiniMax): API pricing, context window, capabilities and uptime, from the same data as each model page, at the WandAPI default channel price. Input costs 1.26 Credits / M for MiniMax M2.5 and 1.26 Credits / M for MiniMax M3. Context windows are 200K and 1M.
MiniMax M2.5minimax-m2.5MiniMax M3minimax-m3Specs side by side
| Item | MiniMax M2.5minimax-m2.5 | MiniMax M3minimax-m3 |
|---|---|---|
| Input (prompt) · Credits / M | 1.26≈ $0.191 | 1.26≈ $0.191 |
| Output (completion) · Credits / M | 5.04≈ $0.766 | 5.04≈ $0.766 |
| Cache read · Credits / M | 0.126 | 0.252 |
| Context window | 200K | 1M |
| Max output | 128K | 128K |
| 7-day uptime | — | 70.8% |
| Latest probe latency | — | 2.5s |
| Measured throughput | 4.5 t/s | 7.0 t/s |
| Protocols | Anthropic Messages | Anthropic Messages |
| Sample monthly cost (10M in + 2M out) | 22.68≈ $3.45 | 22.68≈ $3.45 |
Prices are the WandAPI default channel price in Credits (1 Credit = ¥1); green marks the better side. Uptime and latency come from WandAPI probes; a dash means the model is not probed yet.
Capabilities
FAQ
Which has the larger context window, MiniMax M2.5 or MiniMax M3?
MiniMax M2.5 offers 200K tokens, MiniMax M3 offers 1M tokens.
Can I switch between MiniMax M2.5 and MiniMax M3 at any time?
Yes. WandAPI calls every model with the same API key; switching is just changing the model field in the request, and billing is split per model in one console.
How do the capabilities of MiniMax M2.5 and MiniMax M3 differ?
Only MiniMax M3: Vision, Prompt Caching, Video. Every other capability is either supported by both or by neither; see the capability table above.
More comparisons
One API key for both MiniMax M2.5 and MiniMax M3
Free credits on sign-up, pay as you go, no monthly fee.