Tools
Inworld Tts
By Unofficial OpenClaw speech provider plugin for Inworld TTS
Install
npm install
```Configuration Example
{
"method": "talk.speak",
"params": {
"text": "こんにちは、OpenClawです。",
"voiceId": "alternate-voice-id",
"modelId": "inworld-tts-1.5-max",
"speed": 0.95,
"normalize": "on"
}
}README
# Inworld TTS for OpenClaw
Unofficial, community-maintained OpenClaw plugin that adds the Inworld TTS provider.
It is built for OpenClaw's extension loader and ClawHub publication, not as a generic Inworld SDK wrapper.
- Package name: `openclaw-inworld-tts`
- Extension/plugin id: `inworld-tts`
- Speech provider id: `inworld`
- License: `MIT`
- Required auth: `INWORLD_API_KEY` or `messages.tts.providers.inworld.apiKey`
The extension registers the `inworld` speech provider, so OpenClaw config still uses `messages.tts.provider: "inworld"` and `messages.tts.providers.inworld`.
## Requirements
- OpenClaw `>=2026.4.1`
- Inworld API key via `INWORLD_API_KEY` or `messages.tts.providers.inworld.apiKey`
- An Inworld `voiceId`
## Install
Once published, install it like any other external OpenClaw plugin:
```bash
openclaw plugins install openclaw-inworld-tts
```
OpenClaw checks ClawHub first for bare npm-safe package names and falls back to npm if needed.
## Minimal TTS config
```json5
{
messages: {
tts: {
provider: "inworld",
providers: {
inworld: {
voiceId: "your-inworld-voice-id",
},
},
},
},
}
```
## Example with all supported settings
```json5
{
messages: {
tts: {
provider: "inworld",
providers: {
inworld: {
apiKey: "optional-if-INWORLD_API_KEY-is-set",
baseUrl: "https://api.inworld.ai",
voiceId: "your-inworld-voice-id",
modelId: "inworld-tts-1.5-max",
temperature: 0.7,
timestampType: "WORD",
applyTextNormalization: "auto",
languages: ["ja-JP", "en-US"],
audioConfig: {
audioEncoding: "MP3",
sampleRateHertz: 44100,
bitRate: 128000,
speakingRate: 1,
},
},
},
},
},
}
```
## Talk mode example
Talk mode uses `talk.provider` and `talk.providers.<provider>`.
For Inworld, keep the Talk provider name as `inworld`.
```json5
{
talk: {
provider: "inworld",
providers: {
inworld: {
apiKey: "optional-if-INWORLD_API_KEY-is-set",
baseUrl: "https://api.inworld.ai",
voiceId: "your-inworld-voice-id",
modelId: "inworld-tts-1.5-max",
temperature: 0.7,
applyTextNormalization: "auto",
audioConfig: {
audioEncoding: "MP3",
sampleRateHertz: 44100,
bitRate: 128000,
speakingRate: 1,
},
},
},
},
}
```
### `talk.speak` override example
The public `talk.speak` gateway method accepts the generic Talk override fields.
For the Inworld provider, the useful per-request overrides are `voiceId`, `modelId`, `speed`, and `normalize`.
```json
{
"method": "talk.speak",
"params": {
"text": "こんにちは、OpenClawです。",
"voiceId": "alternate-voice-id",
"modelId": "inworld-tts-1.5-max",
"speed": 0.95,
"normalize": "on"
}
}
```
Provider-specific defaults still belong in `talk.providers.inworld`.
## Discord voice example
Discord voice playback can override the global `messages.tts` config with `channels.discord.voice.tts`.
```json5
{
channels: {
discord: {
voice: {
enabled: true,
tts: {
provider: "inworld",
providers: {
inworld: {
voiceId: "your-inworld-voice-id",
modelId: "inworld-tts-1.5-max",
audioConfig: {
audioEncoding: "OGG_OPUS",
sampleRateHertz: 48000,
},
},
},
},
},
},
},
}
```
You can also override per Discord account with `channels.discord.accounts.<account-id>.voice.tts` when you want one bot account to speak differently from the global Discord voice default.
## Config reference
| Key | Type | Default | Notes |
| ----------------------------- | ---------- | ------------------------ | ------------------------------------------------------------------------------------------------- |
| `apiKey` | `string` | unset | Falls back to `INWORLD_API_KEY`. Required for voice listing and synthesis if env is not set. |
| `baseUrl` | `string` | `https://api.inworld.ai` | Trailing slash is removed automatically. |
| `voiceId` | `string` | unset | Required for synthesis. |
| `modelId` | `string` | `inworld-tts-1.5-max` | Supported models are listed below. |
| `temperature` | `number` | provider default | Allowed range is `> 0` and `<= 2`. `0` is treated as unset and omitted from the API request. |
| `timestampType` | `string` | unset | Allowed values are `TIMESTAMP_TYPE_UNSPECIFIED`, `WORD`, `CHARACTER`. |
| `applyTextNormalization` | `string` | unset | Allowed values are `auto`, `on`, `off`. `auto` maps to Inworld's unspecified/default behavior. |
| `languages` | `string[]` | unset | Used when listing voices. Language values are normalized to Inworld locale style such as `JA_JP`. |
| `audioConfig.audioEncoding` | `string` | target-dependent | Allowed values are listed below. |
| `audioConfig.sampleRateHertz` | `number` | target-dependent | Allowed values are listed below. |
| `audioConfig.bitRate` | `number` | target-dependent | Must be a positive integer when set. |
| `audioConfig.speakingRate` | `number` | unset | Allowed range is `0.5` to `1.5`. |
## Supported values
### `modelId`
- `inworld-tts-1`
- `inworld-tts-1-max`
- `inworld-tts-1.5-mini`
- `inworld-tts-1.5-max`
### `audioConfig.audioEncoding`
- `LINEAR16`
- `MP3`
- `OGG_OPUS`
- `ALAW`
- `MULAW`
- `FLAC`
- `PCM`
- `WAV`
### `audioConfig.sampleRateHertz`
- `8000`
- `16000`
- `22050`
- `24000`
- `32000`
- `44100`
- `48000`
## Target-specific defaults
When `audioConfig` is omitted, OpenClaw picks defaults based on the output target.
| Target | `audioEncoding` | `sampleRateHertz` | `bitRate` | Notes |
| -------------------------- | --------------- | ----------------- | --------- | -------------------------------------------------------------------- |
| Normal reply / file output | `MP3` | `44100` | `128000` | Default for general outbound audio. |
| Voice note | `OGG_OPUS` | `48000` | unset | Marked as voice-compatible by the provider. |
| Telephony | `LINEAR16` | `8000` | unset | Forced for telephony output regardless of the normal reply defaults. |
## OpenClaw behavior notes
- The extension registers only a speech provider. It does not add a new channel or realtime voice transport by itself.
- `messages.tts.provider` must remain `inworld`. The plugin name `inworld-tts` is only for OpenClaw's plugin loader.
- `messages.tts.providers.inworld.languages` affects `listVoices`, not the synthesis request.
- `talk.providers.inworld` can store provider-specific Talk defaults such as `baseUrl`, `temperature`, `applyTextNormalization`, and `audioConfig`.
- Public `talk.speak` requests use the generic Talk schema; for Inworld the practical per-request overrides are `voiceId`, `modelId`, `speed`, and `normalize`.
- `channels.discord.voice.tts` overrides `messages.tts` for Discord voice playback only.
- Telephony synthesis forces `LINEAR16` at `8000` Hz.
## Service docs
- [Inworld API introduction](https://docs.inworld.ai/api-reference/introduction)
- [Inworld TTS synthesize speech](https://docs.inworld.ai/api-reference/ttsAPI/texttospeech/synthesize-speech)
- [Inworld list voices](https://docs.inworld.ai/api-reference/voiceAPI/voiceservice/list-voices)
## ClawHub packaging notes
- `package.json` includes `openclaw.compat` and `openclaw.build`, which are required for ClawHub-published external plugins.
- `package.json` keeps `openclaw` as a peer dependency because the host runtime provides it.
- The local test setup uses lightweight SDK shims so `pnpm test` and `pnpm check` can run without a sibling OpenClaw checkout.
## Validation
From this package directory, first install dependencies so `pnpm-lock.yaml` is generated locally:
```bash
pnpm install
```
Then run:
```bash
pnpm test
pnpm check
```
- `pnpm test` runs the local extension Vitest suite, including `readme.test.ts`.
- `pnpm check` runs the local formatting check, local lint, and then the local extension test suite.
## Support
[](https://github.com/sponsors/livingghost)
If this plugin is useful, you can support development here:
- [Sponsor on GitHub](https://github.com/sponsors/livingghost)
tools
Comments
Sign in to leave a comment