Doubao cloud TTS: API key and setup

Doubao cloud TTS turns book text into speech. This guide uses Origo X's built-in Doubao preset for speech synthesis 2.0. Activate the service in your own Volcengine account, obtain an API key, and enter it in the app.

Official application links

Get or manage a Doubao Speech API key: sign in to Volcengine to open key management.
Official quick start for the new console: registration, identity verification and service activation (Chinese).
Official API key instructions: obtaining, disabling and deleting keys.

Use the new Doubao Speech console. An ordinary Volcengine Ark chat-model key, an AccessKey / SecretKey pair, or an AppID / Access Token from the legacy speech console cannot be pasted directly into the API key field for this setup.

1. Register and activate speech synthesis

Open the official application link and register or sign in. Complete the identity verification requested for first-time users.
Enter the new Doubao Speech console and check the selected project. You can start with the default project.
Under service activation (开通管理), enable Doubao speech synthesis 2.0 (豆包语音合成模型 2.0) and confirm it is available. If you use the official one-click activation option, check the synthesis 2.0 status afterward.
Review your trial allowance, expiry and billing terms before continuing. Having a key does not by itself mean the service is active.

Trial allowances depend on your account. An activated service may continue with postpaid billing after its trial allowance runs out. Review the official quick start. Origo X account benefits do not include Volcengine speech usage charges.

2. Obtain your API key

Open API key management in the project where you activated the service. This link selects default; switch projects if needed.
Obtain an available key. To create one, use the console's creation option and follow its prompts; a name such as Origo X listening helps identify it.
Copy the complete key. Do not include quotes, Bearer, or X-Api-Key:.

The new console uses a single API key for authentication. The built-in preset does not need a separate AppID; see the official key instructions.

3. Enter the key in Origo X

If Cloud TTS is missing, check downloads and updates. Follow any app prompt to sign in or enable the required account access before continuing.

Open Settings → Cloud TTS.
Under Providers, select 豆包 (Doubao).
Set Speech model to 豆包语音合成 2.0 and start with the Vivi 2.0 voice.
Paste your key into API Key.
Select Preview voice and confirm that the sample plays. Preview pauses book playback and may incur provider charges.
Select Save and use to save the configuration and select cloud speech.

The preset already supplies the endpoint and model. To check them, open Advanced customization:

Setting	Value for this guide
Provider	豆包 (Doubao)
Speech model	Doubao speech synthesis 2.0
Example voice	Vivi 2.0
Audio format	`mp3`
API Key	Your complete key from the new Doubao Speech console

In Advanced customization, Custom model ID is seed-tts-2.0 and the example Custom voice ID is zh_female_vv_uranus_bigtts. Keep Service URL set to the complete address below:

https://openspeech.bytedance.com/api/v3/tts/unidirectional

The endpoint and resource ID match the official HTTP synthesis documentation. Keep the preset for your first setup. Other voices must match your model version and account permissions.

4. Start listening

Open a book and its read-aloud player. In listening settings, select Cloud TTS and your saved Doubao configuration, then start playback. Try a short passage to check the voice, speed and connection.

You can also open the configuration through Set up cloud voice or Voice and connection settings in the player. See Read aloud for playback controls and the sleep timer.

Troubleshooting

Key, request or synthesis errors

Check that the key is complete, active and from the new Doubao Speech console. In the same project, check synthesis 2.0 activation, remaining allowance or balance, and voice permissions. Then check network access to the official endpoint and retry the preview.

You still hear the system voice

Confirm that Cloud TTS and your saved Doubao configuration are selected. If Use system voice on failure is enabled, a failed cloud request may switch to system speech. Use Preview voice in the editor to diagnose the cloud configuration; preview reports cloud errors directly.

Only one voice fails

Try Vivi 2.0 again and check voice permissions in the console. seed-tts-2.0 uses 2.0 voices. Do not mix voice IDs from other model versions. You can enter other official voice IDs under Advanced customization.

The key field is empty when editing or changing devices

If a key is already saved, the editor may indicate this while leaving the input empty. A blank input retains the saved key; paste a new one to replace it. On a new device, check whether a key is available and enter it again if needed.

Your key was exposed

Disable or delete it in official key management, obtain a replacement, and update and preview it in the app. For support, provide the app version, operating system, model, voice and error message with the key removed. Keep the complete key out of screenshots.

Official links and instructions checked on 11 October 2026. Console labels, activation options and allowances may change; follow the current official prompts when applying.