minimax-speech
Use this skill when producing speech, narration, dubbing, localization, advertising voice, voice-clone previews, or interactive voice with MiniMax speech/audio APIs. It covers MiniMax T2A HTTP, WebSocket streaming, async long-form TTS, Speech 2.8/2.6/02 model selection, system and custom voices, rapid voice cloning, voice design, pronunciation/language/emotion controls, pricing and limits, artifact custody, consent and rights checks, and production QA.
Profile →npx skills add calesthio/generative-media-skills --skill minimax-speech --agent claude-code
Same command for any agent — swap --agent for codex, cursor, copilot.
Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.
What it does
The skill directs the agent to generate speech assets for narration, localization, ads, voice cloning previews, and interactive voice using MiniMax APIs. It covers text-to-audio via HTTP and WebSocket, asynchronous long-form TTS, voice management (system, cloned, and designed voices), rapid voice cloning, voice design, subtitle timing, and an ASR-based validation check inside cloning. It also defines model choices, endpoint usage, and production QA steps.
How it works
- It enumerates the supported MiniMax capabilities: HTTP T2A, WebSocket streaming, async T2A, and voice management operations. It lists cloning, voice design, and subtitle/timestamping workflows.
- It prescribes model selection by using the newest supported model unless constrained, with specific model IDs for different use cases (e.g., speech-2.8-hd for high-quality final narration, speech-2.8-turbo for interactive/latency-sensitive work, and legacy 2.6/02 variants for compatibility).
- It defines endpoint usage: HTTP for full audio generation within sync limits, WebSocket for streaming or long scripts, and Async for long-form narration up to 50,000 characters, with lifecycle details for task IDs, tokens, and file retrieval.
- It lists core T2A controls (voice_id, speed, vol, pitch, emotion, text normalization, language_boost, pronunciation_dict, inline pronunciation, pause markers, audio settings, subtitle controls, and voice_modify) and warns against overloading controls.
- It provides guidance for script writing specific to MiniMax voices, voice casting workflow, rapid cloning steps, consent requirements, and design considerations for audio provenance and QA.
- It outlines artifact custody, pricing, rate limits, and QA expectations.
When to use it
Use this skill for producing speech-focused content (narration, localization, ads, VO, or voice-clone previews) and for workflows requiring voice design, cloning with consent, or production QA. It is not intended for video, music generation, image generation, or general LLM calls unless they affect speech production.
What it can touch
- Endpoints and controls for MiniMax:
POST /v1/t2a_v2,wss://api.minimax.io/ws/v1/t2a_v2,POST /v1/t2a_async_v2,POST /v1/voice_clone,POST /v1/files/upload,POST /v1/voice_design, and voice management endpoints such asPOST /v1/get_voiceandPOST /v1/delete_voice. - Metadata fields:
voice_setting.*,language_boost,pronunciation_dict.tone,subtitle_enable,subtitle_type,audio_setting.*,pause markers, andvoice_modify. - Cloning workflow, prompt_audio, consent checks, and artifact custody procedures.
Caveats
- The skill notes that model IDs, prices, rate limits, language support, and policy language are volatile as of 2026-07-10; re-check before paid or regulated production.
- Standalone transcription, full dubbing, or voice conversion beyond T2A/P2A are not documented as standalone production APIs and must be combined with other providers or clarified.
- Availability and behavior of certain features (e.g., fluent/whisper emotions, 2.8 vs 2.6 model capabilities) vary by model and endpoint; verify exact parameter names and supported options for the endpoint used.
- Cloning requires explicit consent and proper governance; undocumented automatic persistence or reuse policies may apply to cloned voices.
# MiniMax speech production Use MiniMax when the production goal is generated speech: narration, localized VO, ad reads, audiobook-style segments, synthetic dialogue, product walkthroughs, voice-agent responses, or approved voice cloning. Do not use this skill for MiniMax video, music generation, image generation, or general LLM calls except where they affect speech production. The facts below wer
What does the minimax-speech skill do?
Use this skill when producing speech, narration, dubbing, localization, advertising voice, voice-clone previews, or interactive voice with MiniMax speech/audio APIs. It covers MiniMax T2A HTTP, WebSocket streaming, async long-form TTS, Speech 2.8/2.6/02 model selection, system and custom voices, rapid voice cloning, voice design, pronunciation/language/emotion controls, pricing and limits, artifact custody, consent and rights checks, and production QA.
How do I install it?
Run `npx skills add calesthio/generative-media-skills --skill minimax-speech --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.
Where does this skill come from?
From calesthio/generative-media-skills, a repository with 112 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.
Is a popular skill a good skill?
Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.