IMA AI Text To Speech — seed-tts, DouBao

核心功能：基于Seed TTS 2.0模型，将文本、脚本转为自然语音，支持选择多音色，调节情绪、语速、音量，可记忆用户偏好。典型场景：视频旁白、产品演示、讲解类内容、社交平台配音等。输入为待朗读文本，可补充指定音色等参数；输出为语音链接直接推送，需配置IMA API密钥，支持命令行调用。

allenfancy-gan

热度 65.14

最近更新 2026-05-11 17:29:44

分类：创作依赖：代码执行 / 第三方 API / 读写文件应用场景：视频编辑 / 社交媒体 / 故事创作

安装技能

复制下面的话术发给你的 Agent，它会通过 secure-skills 检索并安装该技能。

执行命令安装此 skill： npx secure-skills-cli add ima-tts-ai

Metadata

name: IMA TTS Generator
version: 1.0.0
category: file-generation
author: IMA Studio (imastudio.com)
keywords: imastudio, tts, text-to-speech, speech synthesis, voice, voiceover, narration, 语音合成, 文字转语音, IMA, TTS, AI配音, 朗读, 播报, seed-tts, seed-tts-2.0, dubbing, audiobook
argument-hint: [text to speak or 要朗读的文本]
description: Convert text, scripts, and captions into natural voiceovers for videos, explainers, product demos, and social posts.
requires: {"env":["IMA_API_KEY"],"runtime":["python3"],"packages":["requests"],"primaryCredential":"IMA_API_KEY"}
metadata: {"openclaw":{"primaryEnv":"IMA_API_KEY","homepage":"https://imastudio.com","requires":{"bins":["python3"],"env":["IMA_API_KEY"]}}}
persistence: {"readWrite":["~/.openclaw/memory/ima_prefs.json","~/.openclaw/logs/ima_skills/"],"retention":"Logs are auto-cleaned after 7 days; preferences remain until user deletes them."}

IMA TTS AI — Text-to-Speech Generator

For complete API documentation, security details, all parameters, speaker list, and Python examples, read SKILL-DETAIL.md.

Model ID Reference (CRITICAL)

Friendly Name	model_id	Notes
Seed TTS 2.0	`seed-tts-2.0`	✅ Default and only supported model

Sub-models (via extra-params):

seed-tts-2.0-expressive — More expressive, emotional (default)
seed-tts-2.0-standard — More stable, neutral

When User Says "帮我制作旁白/配音"

Must ask first:

Question	Parameter	Required
要朗读的内容/文案	`prompt`	✅ Yes

Recommend asking:

Question	Parameter	Options
音色/发音人	`speaker`	魅力苏菲、Vivi、云舟、大壹等 (see SKILL-DETAIL.md)

Optional:

Question	Parameter	Range
情感/情绪	`audio_params.emotion`	neutral, sad, angry
语速	`audio_params.speech_rate`	[-50, 100], 0=normal
音量	`audio_params.loudness_rate`	[-50, 100], 0=normal

User Input Parsing

User says	Parameter	Value
旁白/配音/朗读	prompt + speaker	Ask for content first
女声/female	speaker	e.g. `zh_female_vv_uranus_bigtts`
男声/male	speaker	e.g. `zh_male_sophie_uranus_bigtts`
语速快/slow	audio_params.speech_rate	Positive/negative value
expressive/standard	model	Sub-model selection

Script Usage

# List available TTS models
python3 {baseDir}/scripts/ima_tts_create.py --api-key $IMA_API_KEY --list-models

# Generate speech (default model: seed-tts-2.0)
python3 {baseDir}/scripts/ima_tts_create.py \
  --api-key $IMA_API_KEY \
  --model-id seed-tts-2.0 \
  --prompt "Text to be spoken here." \
  --user-id {user_id} \
  --output-json

# With speaker and emotion
python3 {baseDir}/scripts/ima_tts_create.py \
  --api-key $IMA_API_KEY \
  --model-id seed-tts-2.0 \
  --prompt "阳光青年音色测试，你好世界。" \
  --extra-params '{"model":"seed-tts-2.0-expressive","speaker":"zh_male_sophie_uranus_bigtts","audio_params":{"emotion":"neutral"}}' \
  --user-id {user_id} \
  --output-json

Sending Results to User

# ✅ CORRECT: Use remote URL directly
message(action="send", media=audio_url, caption="✅ 语音合成成功！\n• 模型：[Name]\n• 耗时：[X]s\n• 积分：[N pts]\n\n🔗 原始链接：[url]")

# ❌ WRONG: Never download to local file

UX Protocol (Brief)

Pre-generation: "🔊 开始语音合成… 模型：[Name]，预计[X~Y]秒，消耗[N]积分"
Progress: Every 10-15s: "⏳ 语音合成中… [P]%"
Success: Send audio via media=audio_url + include link in caption
Failure: Natural language error + suggest retry. See SKILL-DETAIL.md for error translation.

Never say to users: script names, API endpoints, attribute_id, technical parameter names.

Environment

Base URL: https://api.imastudio.com Headers: Authorization: Bearer $IMA_API_KEY · x-app-source: ima_skills · x_app_language: en

Core Flow

GET /open/v1/product/list?app=ima&platform=web&category=text_to_speech → get attribute_id, credit
POST /open/v1/tasks/create → get task_id
POST /open/v1/tasks/detail → poll every 2-5s until resource_status==1

MANDATORY: Always query product list first. attribute_id is required.

Estimated Generation Time

Model	Estimated Time	Poll Every
seed-tts-2.0	5~30s	3s

User Preference Memory

Storage: ~/.openclaw/memory/ima_prefs.json

Save when user explicitly says "用XXX音色" / "默认用XXX"
Clear when user says "换个音色" / "推荐一个"

Popular Speakers (Quick Reference)

Category	Speaker Name	speaker ID
通用	魅力苏菲	`zh_male_sophie_uranus_bigtts`
通用	Vivi	`zh_female_vv_uranus_bigtts`
通用	云舟	`zh_male_m191_uranus_bigtts`
视频配音	大壹	`zh_male_dayi_uranus_bigtts`
角色扮演	知性灿灿	`zh_female_cancan_uranus_bigtts`

Full speaker list: See volcengine_tts_timbre_list.json in project or SKILL-DETAIL.md.

⚠️ Important: Use native format (*_uranus_bigtts), NOT BV*_streaming format.