The cheapest any-modality model on this site — text, images, audio, and video in for $0.12 per million input tokens. Real rates for both MiMo models, a workload calculator, and the budget comparisons.
| MODEL | INPUT $/1M | CACHED IN | OUTPUT $/1M | BLENDED | CONTEXT | YOUR COST |
|---|---|---|---|---|---|---|
MiMo-V2.5-Pro xiaomi/mimo-v2.5-pro · via OpenRouter | $0.30 | — | $0.61 | $0.377 | 1.05M | $0.60 |
MiMo-V2.5 xiaomi/mimo-v2.5 · via OpenRouter | $0.12 | — | $0.24 | $0.15 | 1M | $0.24 |
USD per 1M tokens from Xiaomi’s official pricing page as of August 31, 2026. Blended assumes a 3:1 input:output ratio.
MiMo-V2.5 takes text, images, audio, and video — the full input surface — for $0.12 input / $0.24 output per 1M tokens with a 1M context. No other model in our dataset offers audio and video input anywhere near this price; for voice and media pipelines on a budget it is essentially the default answer.
The trap is the naming: MiMo-V2.5-Pro, at $0.30 / $0.61, is the stronger text model but drops every modality except text. If you picked Pro assuming it does everything the base model does plus more, your image inputs will bounce.
Xiaomi does not run a first-party USD developer platform, so these are OpenRouter serve rates (open weights mean you can also self-host). Rates on OpenRouter can vary slightly by underlying host — check the model page before committing volume.
Via OpenRouter: MiMo-V2.5 costs $0.12 input / $0.24 output per 1M tokens with text, image, audio, and video input; MiMo-V2.5-Pro costs $0.30 / $0.61 and is text-only. Both ship open weights.
Xiaomi publishes no first-party USD rate card, so the practical way to use MiMo over an API is through OpenRouter (or by self-hosting the open weights). The prices here are the OpenRouter serve rates on our verification date and can vary slightly by host.
V2.5 is omnimodal — text, images, audio, and video in — while V2.5-Pro is a stronger but text-only model at about 2.5× the price. Media pipelines want V2.5; hard text tasks on a budget want Pro.
On price per modality nothing touches V2.5: $0.15/1M blended with audio and video input included, versus $0.85 for Gemini Flash-Lite (the cheapest closed omnimodal option). For text-only work, GLM-5.3-Flash and Nemotron 3 Super compete at similar or lower blended prices.
Also see: OpenAI API pricing · Claude API pricing · Gemini API pricing · Grok API pricing · DeepSeek API pricing · Kimi API pricing · GLM API pricing · MiniMax API pricing · all 58 models
ReqKey gives every key a credit balance — validate, deduct, enforce limits, and watch per-consumer MiMo spend in real time.