DIGIT ElevenLabs Voice Selector

The name is a lie, and it's a good lie. This isn't a node that talks to ElevenLabs - it's a dropdown that resolves a human-friendly voice name to the opaque voice ID ElevenLabs actually wants. "Sarah (female, american)" becomes 21m00Tcm4TlvDq8ikWAM (or whatever the real ID is) with zero API calls, zero keys, and zero cost. It exists purely to save you from memorizing or copy-pasting UUIDs, and it does that job so well you'll forget it's there.

That's the whole utility: the ElevenLabs suite needs a voice_id string on every Text to Speech node, and raw IDs are the least memorable things in the world. This node turns the ID into a name you can actually navigate. It's a string mapper, not an API client.

How it works

Under the hood there's a hardcoded table in the pack: 21 predefined ElevenLabs voices, each mapped to its ID. You pick voice from the dropdown - Roger, Sarah, Laura, Charlie, George, Callum, River, Harry, Liam, Alice, Matilda, Will, and the rest - and the node returns the matching voice_id on its single STRING output.

The one input:

• voice - an enum of 21 named voices, each labeled with gender and accent (male/female/neutral plus american/british/australian).

The one output:

• voice_id - the ID string. Wire it into Text to Speech's voice_id input and you're done. No API key, no network, no auth - you could use this node with the whole internet down.

Install

It's one node in the comfyui-digit pack, so you get it with the family. ComfyUI Manager → search comfyui-digit → install → restart, or:

cd ComfyUI/custom_nodes
git clone https://github.com/thedepartmentofexternalservices/comfyui-digit.git
cd comfyui-digit
pip install -r requirements.txt

Restart ComfyUI. It lives under DIGIT/ElevenLabs.

Common issues

Honestly, the failure modes are close to zero - this node doesn't touch the network or need a key. The only thing to keep in mind is that it's a fixed list of predefined voices: the 21 built-in ElevenLabs voices, not your account's custom or cloned voices. If you clone a voice with the Voice Clone node, that voice_id comes straight off the clone node's output - you don't route it through this selector, because it won't be in the list. Use this one for stock voices, the clone node for yours.

One more thing: the display name on the dropdown doesn't guarantee the voice stays in ElevenLabs' public lineup forever. If you build a workflow around a specific name and it eventually vanishes upstream, the node still resolves to the old ID and the API will error. That's a slowly-rotting edge case, not something you'll hit this week.