[Integrations](https://spiral.ai/integrations) › [opencode](https://spiral.ai/integrations/opencode) › [Qwen3-TTS 0.6B Clone Voice](https://spiral.ai/integrations/opencode/qwen3-tts-0-6b-clone-voice)

# How to Use Qwen3-TTS 0.6B Clone Voice in opencode

Qwen3-TTS 0.6B Clone Voice turns a recording into a reusable voice profile for speech generation. It uses the smaller Base model compared with Qwen3-TTS 1.7B Clone Voice, whose underlying model scored better on Qwen's English and Chinese speech accuracy tests.

|  |  |
| --- | --- |
| Model | Qwen3-TTS 0.6B Clone Voice |
| Available through | fal |
| Category | Audio |
| Price | Up to $0.00091 per run at the default inputs for 60 seconds of input. |
| Returns | audio |
| Runs | In the background |
| Setup | About 2 minutes, no API key |

## Prompts to try with Qwen3-TTS 0.6B Clone Voice in opencode

- Use Qwen3-TTS 0.6B Clone Voice on the clean presenter recording in this project's audio assets. Save the returned voice profile beside its matching transcript for future tutorial narration.
- Use Qwen3-TTS 0.6B Clone Voice to extract a profile from the approved narrator sample in this repository. Check that the recording is under five minutes and preserve its exact transcript.
- Use Qwen3-TTS 0.6B Clone Voice on each approved character recording in the project. Store a separate profile and reference transcript for each character.
- Use Qwen3-TTS 0.6B Clone Voice to prepare profiles from the three short presenter takes in our audio folder. Label each result by its source take so I can compare them in a later speech test.
- Use Qwen3-TTS 0.6B Clone Voice to prepare the instructor profile for our course narration workflow. Use the existing clean reference clip and keep its spoken transcript with the result.

## How to set up Qwen3-TTS 0.6B Clone Voice in opencode

Pick the way that suits you. Each takes about two minutes.

### Use Spiral's MCP server

1. **Add the server**: Leave out `--global` to add Spiral to the current project only. No header and no client ID or secret: opencode registers itself with Spiral.

```bash
opencode mcp add spiral --global --url https://api.spiral.ai/mcp
```

2. **Sign in**: Your browser opens. Sign in to Spiral, pick the organization and click Authorize, then return to the terminal. No API key. Inside opencode, `/mcps` starts the same sign-in.

```bash
opencode mcp auth spiral
```

3. **Check it**: Spiral shows as connected.

```bash
opencode mcp list
```

### Use the Spiral skill

1. **Send opencode this message**: Your agent reads Spiral's skill and sets itself up.

```
set up https://spiral.ai/SKILL.md
```

2. **Sign in or paste a key**: Approve Spiral in the browser, or give the agent a key from the console.

### Use the Spiral CLI

1. **Install and sign in**: Installs the Spiral CLI, signs you in and adds its skill to your coding agents.

```bash
npx -y spiral-cli@latest init
```

### Call the Spiral API

1. **Create an API key**: In the console, under API keys.

2. **Run Qwen3-TTS 0.6B Clone Voice**: One request; the result comes back by polling or a webhook.

```bash
curl https://api.spiral.ai/v1/runs \
  -H "Authorization: Bearer $SPIRAL_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: $(uuidgen)" \
  -d '{
    "provider": "fal",
    "model": "fal-ai/qwen-3-tts/clone-voice/0.6b",
    "input": {
      "audio_url": "https://storage.googleapis.com/falserverless/example_inputs/qwen3-tts/clone_in.mp3"
    }
  }'
```

Then paste one of the prompts above: opencode finds Qwen3-TTS 0.6B Clone Voice on Spiral, shows the price and runs it.

Setup guide in the docs: https://spiral.ai/docs/quickstarts/opencode

## Qwen3-TTS 0.6B Clone Voice tools and prices

- [Qwen 3 TTS 0.6B · Clone-Voice](https://spiral.ai/tools/fal/fal-ai/qwen-3-tts/clone-voice/0.6b) (fal): $0.00091 per minute.

## Settings you can ask for

| Setting | Allowed values | Default | Description |
| --- | --- | --- | --- |
| reference_text |  | — | Optional reference text that was used when creating the speaker embedding. Providing this can improve synthesis quality when using a cloned voice. |

## What can you do with Qwen3-TTS 0.6B Clone Voice in opencode?

Prepare reusable narrator and character voice profiles from recordings in your project. Keep each profile linked to its reference transcript so later speech generation work can reuse the same source.

### Product tutorial narration

Extract a presenter's voice profile from a clean recording, then use it in a speech generation workflow for tutorial updates.

### Recurring audio briefings

Prepare a reusable narrator profile for daily or weekly briefings. Keep voice extraction separate from each briefing's script.

### Course narration

Create an instructor voice profile for use when synthesizing lesson introductions, summaries, and revised explanations.

### Game character voice preparation

Extract separate profiles from recorded character performances, then use those profiles when producing dialogue with a compatible speech model.

### Multilingual narration preparation

Prepare a speaker profile for a narration workflow covering supported languages such as English, Spanish, French, and Japanese. Review each generated language separately.

### Voice reference comparisons

Extract profiles from several clean recordings of the same speaker, then compare speech generated from the same test script before choosing a reference.

## Overview

- [fal](https://spiral.ai/providers/fal)
- [All opencode guides](https://spiral.ai/integrations/opencode)

## Qwen3-TTS 0.6B Clone Voice in other agents

- [Amp](https://spiral.ai/integrations/amp/qwen3-tts-0-6b-clone-voice)
- [ChatGPT](https://spiral.ai/integrations/chatgpt/qwen3-tts-0-6b-clone-voice)
- [Claude](https://spiral.ai/integrations/claude-desktop/qwen3-tts-0-6b-clone-voice)
- [Claude Code](https://spiral.ai/integrations/claude-code/qwen3-tts-0-6b-clone-voice)
- [Cline](https://spiral.ai/integrations/cline/qwen3-tts-0-6b-clone-voice)
- [Codex](https://spiral.ai/integrations/codex/qwen3-tts-0-6b-clone-voice)
- [Cursor](https://spiral.ai/integrations/cursor/qwen3-tts-0-6b-clone-voice)
- [Devin](https://spiral.ai/integrations/devin/qwen3-tts-0-6b-clone-voice)
- [dots](https://spiral.ai/integrations/dots/qwen3-tts-0-6b-clone-voice)
- [Factory Droid](https://spiral.ai/integrations/factory-droid/qwen3-tts-0-6b-clone-voice)
- [Gemini](https://spiral.ai/integrations/gemini/qwen3-tts-0-6b-clone-voice)
- [Gemini CLI](https://spiral.ai/integrations/gemini-cli/qwen3-tts-0-6b-clone-voice)
- [Google Antigravity](https://spiral.ai/integrations/antigravity/qwen3-tts-0-6b-clone-voice)
- [goose](https://spiral.ai/integrations/goose/qwen3-tts-0-6b-clone-voice)
- [Grok](https://spiral.ai/integrations/grok/qwen3-tts-0-6b-clone-voice)
- [Grok Build](https://spiral.ai/integrations/grok-build/qwen3-tts-0-6b-clone-voice)
- [Hermes](https://spiral.ai/integrations/hermes/qwen3-tts-0-6b-clone-voice)
- [Instinct](https://spiral.ai/integrations/instinct/qwen3-tts-0-6b-clone-voice)
- [Manus](https://spiral.ai/integrations/manus/qwen3-tts-0-6b-clone-voice)
- [Muse](https://spiral.ai/integrations/muse/qwen3-tts-0-6b-clone-voice)
- [Muse Code](https://spiral.ai/integrations/muse-code/qwen3-tts-0-6b-clone-voice)
- [n8n](https://spiral.ai/integrations/n8n/qwen3-tts-0-6b-clone-voice)
- [Notion](https://spiral.ai/integrations/notion/qwen3-tts-0-6b-clone-voice)
- [OpenClaw](https://spiral.ai/integrations/openclaw/qwen3-tts-0-6b-clone-voice)
- [Perplexity](https://spiral.ai/integrations/perplexity/qwen3-tts-0-6b-clone-voice)
- [Raycast](https://spiral.ai/integrations/raycast/qwen3-tts-0-6b-clone-voice)
- [VS Code](https://spiral.ai/integrations/vscode/qwen3-tts-0-6b-clone-voice)
- [Windsurf](https://spiral.ai/integrations/windsurf/qwen3-tts-0-6b-clone-voice)
- [Zapier](https://spiral.ai/integrations/zapier/qwen3-tts-0-6b-clone-voice)

## More opencode guides

- [Qwen3 TTS 1.7B Clone Voice](https://spiral.ai/integrations/opencode/qwen3-tts-1-7b-clone-voice)
- [Qwen3-TTS 1.7B](https://spiral.ai/integrations/opencode/qwen3-tts-1-7b)
- [Qwen3 TTS 0.6B](https://spiral.ai/integrations/opencode/qwen3-tts-0-6b)
- [Qwen3-TTS 1.7B Voice Design](https://spiral.ai/integrations/opencode/qwen3-tts-1-7b-voice-design)
- [Qwen Audio 3.0 TTS Flash](https://spiral.ai/integrations/opencode/qwen-audio-3-0-tts-flash)
- [SAM Audio](https://spiral.ai/integrations/opencode/sam-audio)

## Questions about Qwen3-TTS 0.6B Clone Voice in opencode

**Can opencode use Qwen3-TTS 0.6B Clone Voice?**

Yes. Add Spiral to opencode as an MCP server, and opencode can find and run Qwen3-TTS 0.6B Clone Voice like any built-in tool.

**How much does Qwen3-TTS 0.6B Clone Voice cost in opencode?**

Up to $0.00091 per run at the default inputs for 60 seconds of input. You pay from your Spiral balance as you go, with no subscription, and a failed run is not charged. Your opencode plan is separate.

**Do I need a fal account to use Qwen3-TTS 0.6B Clone Voice in opencode?**

No. Spiral runs Qwen3-TTS 0.6B Clone Voice on its own fal account. You sign in to Spiral once and never paste a key.

**How long does it take to set up Qwen3-TTS 0.6B Clone Voice in opencode?**

About two minutes: add the server, then sign in, then check it.

**Does opencode wait while Qwen3-TTS 0.6B Clone Voice runs?**

Qwen3-TTS 0.6B Clone Voice runs in the background. opencode starts the run, checks on it and gives you the audio when it's ready.

**Does Qwen3-TTS Clone Voice produce an audio recording?**

The Clone Voice endpoint produces a reusable voice profile in safetensors format. A compatible speech generation model uses that profile to read new text.

**How much reference speech does Qwen3-TTS need?**

Qwen documents cloning from as little as three seconds of speech. Start with a short, clean recording and test the resulting voice before preparing a larger production.

**How long can the reference recording be?**

The reference recording can be up to five minutes long. A short clip of clean speech is sufficient, so a full interview is unnecessary.

**Which languages does Qwen3-TTS support?**

The underlying model supports Chinese, English, Japanese, Korean, German, French, Russian, Portuguese, Spanish, and Italian. Language selection applies when generating speech with the extracted profile.

**What should the reference transcript contain?**

Use the words spoken in the reference recording. Qwen's implementation can clone from a speaker embedding alone, but its documentation warns that cloning quality may be reduced.

