Low‑latency MiniMax Speech 2.6 Turbo brings multilingual, emotional text-to-speech to Replicate with 300+ voices and real-time friendly pricing
Frequently asked questions
What is MiniMax Speech 2.6 Turbo?
MiniMax Speech 2.6 Turbo is a low-latency, multilingual text-to-speech AI model optimized for real-time applications. It provides 300+ curated voices, supports 40+ languages, offers emotion control, and is available on Replicate via CRAISEE. It is designed for voice bots, chat agents, and interactive UI workflows requiring fast, expressive audio synthesis.
How many languages does MiniMax Speech 2.6 Turbo support?
MiniMax Speech 2.6 Turbo supports 40+ languages, including English, Mandarin Chinese, Cantonese, Japanese, Korean, Spanish, Portuguese, French, German, Arabic, and Hindi. This broad multilingual coverage makes it suitable for global applications requiring natural-sounding voice output across diverse linguistic audiences.
How much does MiniMax Speech 2.6 Turbo cost on Replicate?
MiniMax Speech 2.6 Turbo is priced at $0.06 per 1,000 input tokens on Replicate, with approximately one character equaling one token. This predictable, low-cost pricing makes it accessible for developers building real-time voice applications, interactive agents, or high-volume text-to-speech pipelines without sacrificing quality.
What emotion styles does MiniMax Speech 2.6 Turbo support?
MiniMax Speech 2.6 Turbo supports 10 emotion delivery styles: auto, happy, sad, angry, fearful, disgusted, surprised, calm, fluent, and neutral. These options allow developers and creators to fine-tune the emotional tone of synthesized speech, enabling more expressive and contextually appropriate voice output for diverse use cases.
Can MiniMax Speech 2.6 Turbo clone custom voices?
Yes, MiniMax Speech 2.6 Turbo supports custom cloned voices via compatibility with the minimax/voice-cloning model. In addition to its 300+ built-in system voices, users can supply custom voice IDs to replicate specific speaker characteristics, enabling personalized or branded voice experiences in production applications.
What audio formats does MiniMax Speech 2.6 Turbo output?
MiniMax Speech 2.6 Turbo supports MP3, WAV, FLAC, and raw PCM audio output formats. Sample rates range from 8 kHz to 44.1 kHz, giving developers flexibility to optimize audio quality and file size for different deployment contexts, from telephony systems to high-fidelity web and mobile applications.
Does MiniMax Speech 2.6 Turbo support subtitle or caption generation?
Yes, MiniMax Speech 2.6 Turbo can optionally return sentence-level timestamp metadata alongside the audio output. This subtitle metadata feature supports captioning and accessibility workflows, making it easier to synchronize generated speech with video content or provide text alternatives for hearing-impaired users in compliant applications.
How does MiniMax Speech 2.6 Turbo handle pitch, speed, and volume?
MiniMax Speech 2.6 Turbo allows fine-grained audio control: pitch can be adjusted by semitone offset from −12 to +12, speed can be set from 0.5× to 2.0×, and volume can be tuned on a relative scale of 0 to 10. Inline pause markers like <#0.5#> can also be inserted directly in input text for natural pacing.
