MiniMax · audio
MiniMax Speech 2.8 HD
MiniMax Speech 2.8 HD is a hosted speech-synthesis model. Its documented API rate is $100 per one million characters, with voice controls and synchronous or asynchronous synthesis routes.
Hosted task API
Official sources reviewed 2026-10-02. This page documents access and specifications; it does not report a model benchmark.
Task and deployment contract
Hosted task API; use the exact model ID and the region-specific endpoint in the official documentation.
Exact model ID: speech-2.8-hd. Choose a supported voice_id and audio format from the official API contract.
Speech can be streamed through the documented WebSocket workflow. Measure time to usable audio in your application before making a latency promise.
Character billing is different from text-model token billing. Re-generated narration, pronunciation fixes and separately billed voice creation can change the finished-audio cost.
Preset voices, designed voices and cloned voices use different setup paths. Do not assume a voice ID from another provider transfers to this API.
Billing and limits
$100 / 1M characters
Standard pay-as-you-go T2A model rate. Voice design and cloning are billed separately; consult the official pricing page.
Audio limits are task-specific; no text-token context window is inferred.
Fast Pivot — Access hundreds of AI models from a single API — competitive pricing, reliable uptime, enterprise-grade infrastructure.