MiMo-V2.5-TTS-VoiceClone

MiMo-V2.5-TTS-VoiceClone is a specialized text-to-speech variant within Xiaomi's mimo family that focuses on voice cloning from short audio references. Its intended use is converting written text into natural, fluent speech while letting users configure parameters such as speaking style and target voice. According to the model's listing on DeepInfra, it sits in the Text To Speech category and is described as automatically converting input text into natural and fluent speech output, with the ability to precisely replicate voices from audio samples to enable speech synthesis of any voice.

The model extends the broader MiMo V2.5 family into speech generation, broadening Xiaomi's ecosystem beyond the text-focused general MiMo-V2.5 endpoint that supports multimodal input and very large context windows. For developers, this TTS variant is aimed at use cases such as audiobook narration, personalized voice assistants, and content localization where a consistent synthetic voice is needed. Being available as an open-weights release with no input or output token cost makes it accessible for experimentation and integration into both hobby projects and production pipelines that need voice-cloning capabilities without per-character billing.

Xiaomi Token Plan (Singapore)mimo-v2.5-tts-voiceclonemimo

Open provider page Provider docs

Quick Info

Provider: Xiaomi Token Plan (Singapore)
Model key: mimo-v2.5-tts-voiceclone
Release date: Apr 22, 2026
Last updated: Apr 22, 2026
Input modalities
Output modalities
Capabilities

Cost

A provider subscription or plan supersedes token-based pricing for this model.

Limits

Output tokens: 8,192 tokens
Context window: 8,192 tokens

Latest news about MiMo-V2.5-TTS-VoiceClone

Xiaomi Token Plan (Singapore)

Official sourceDocumentation

The Xiaomi MiMo platform updates page provides first-party release evidence for MiMo-V2.5-TTS-VoiceClone. The 2026-4-23 Token Plan entry lists five new models added to all plans, explicitly naming mimo-v2.5-pro, mimo-v2.5, mimo-v2.5-tts-voiceclone, mimo-v2.5-tts-voicedesign, and mimo-v2.5-tts, and states that tiered to Because the page comes directly from Xiaomi's official MiMo platform changelog, it corroborates that MiMo-V2.5-TTS-VoiceClone is an officially released, commercially available variant on the Xiaomi Token Plan rather than a pre-announcement, and it ties the model's availability to specific subscription tier and discount

Xiaomi Token Plan (Singapore)

Coverage

AIbase's explainer explicitly names MiMo-V2.5-TTS-VoiceClone and describes it as capable of high-fidelity replication of a target voice from a small audio sample while retaining the ability to respond to style instructions and audio tags, with virtual anchors and personalized assistants cited as example scenarios. The The same article documents developer-facing features shared across the series: natural-language instructions that describe desired delivery rather than structured parameters; a layered script mechanism that separately defines character identity, scene atmosphere, and per-line performance; and inline audio tags that can