MiMo V2.5 Pro UltraSpeed

MiMo-V2.5-Pro-UltraSpeed is Xiaomi’s ultra-fast, text-based flagship reasoning model, pairing a trillion-parameter model with a 1,048,576-token context window and generation speeds of roughly 500–1,000 tokens per second. It supports deep thinking, tool calling, streaming output, and context caching, making it well suited to latency-sensitive coding agents, quantitative analysis, real-time risk control, and scientific workflows.

Model overview

Model ID: xiaomi/mimo-v2.5-pro-ultraspeed
Provider: xiaomi
Type: language
Context window: 1,048,576 tokens
Max output tokens: 131,072

Tags: reasoning, tool-use, implicit-caching

Model pricing

Metric	Value
`Input tokens (/1M)`	`$1.30`
`Output tokens (/1M)`	`$2.61`
`Image generation`	`n/a`
`Cached input read (/1M)`	`$0.01`
`Cached input write (/1M)`	`n/a`
`Pricing source`	`gateway`

Related docs

Model catalog

Browse every generated model page.

Open catalog

Pricing matrix

Compare model pricing across the full catalog.

Open pricing matrix