Seed Audio Voice Library

How to Choose a Seed Audio Voice

Start with the language and accent your audience expects. The Seed Audio voice library separates American, British, and Australian English, as well as regional options such as Mexican Spanish and Brazilian Portuguese. Choose the appropriate language filter before comparing individual voices.

Next, consider the use case. The list includes voices categorized for audiobook narration, customer service, education, video voiceover, and role-play. These labels describe the listed scenarios; they do not guarantee a particular performance for every script. Review the supported capabilities alongside the voice name.

To compare candidates, generate the same short passage with each voice and listen for pronunciation, pacing, and delivery. Keep the wording and settings consistent so the comparison focuses on the voice. The Seed Audio prompt guide explains how to describe the surrounding audio scene.

Understanding Voice Capabilities

Emotional expression indicates that the voice supports changes in emotional delivery. Describe the intended feeling in the prompt, such as calm, concerned, or excited, and listen to the result before deciding whether to revise the direction.

Instruction following describes support for explicit delivery directions. For example, a prompt may ask for a measured speaking pace or a particular tone. Keep instructions clear and consistent with the dialogue and scene.

Context following describes support for delivery informed by the surrounding text. Include enough context to explain the speaker’s situation and how the line relates to the rest of the passage.

Not specified means that the supplied listing does not describe additional capabilities for that voice. It does not establish that a capability is supported or unsupported. The voice can still be selected using its “Use this voice” link.

Seed Audio Voice Library FAQ

How do I use a voice from this list?

Select “Use this voice” in the corresponding row. The link opens the homepage audio generator with that voice selected. Enter your audio description, review the settings, and generate when ready. The voice selection is passed to the generation request.

Which languages does Seed Audio support?

Seed Audio supports Chinese, English, Japanese, Korean, Mexican Spanish, Spanish, German, French, Brazilian Portuguese, Thai, Vietnamese, Malay, Filipino, Italian, Russian, Dutch, Polish, and Turkish. Use the language dropdown to browse the voices currently listed for a specific language or accent.

Can I combine a library voice with an audio or image reference?

Choose one mode in the tool: Prompt only, Voice library, Reference audio, or Reference image. Prompt only uses your description without a selected voice or reference. The other modes send only the selected source. To use a listed voice, keep Voice library selected. Read the workspace instructions for the rest of the generation workflow.