Dograh
ElevenLabs voices

Keep your ElevenLabs voices, own the platform

Dograh is the self-hostable voice agent platform with an open-source workflow builder. Bring your ElevenLabs key and voice IDs, cloned voices included, and run them inside a voice agent you self-host or start on Dograh Cloud.

6,000+GitHub starsBSD-2open source$0platform fee, self-hostedAny systemconnects to the tools you use
01 / How it works

Connect the key, pick the voice, call

Example
Add your ElevenLabs keypick a voice IDtest it on a real phone line
01 · Connect

Add ElevenLabs under Providers

BYOKyour own key, your own bill

Add an ElevenLabs connection with your API key, then create a model configuration in BYOK mode.

→ ElevenLabs bills the usage to your account

02 · Voice

Pick the model and the voice

Voice IDany voice your account holds

Choose a model from Flash v2.5 up to Eleven v4 Turbo. Your voices load from the key, and a cloned voice ID can be pasted by hand.

→ Scribe v2 Realtime covers speech-to-text too

03 · Test

Hear it over a phone line

8 kHzhow most phone audio arrives

A browser demo runs at 16 kHz, so listen to the voice on a real call before going live.

→ Record the lines every call needs as clips

02 / What goes where

Your voices in, the platform stays yours

From ElevenLabs
Cloned voice IDsFlash v2.5 to Eleven v4 TurboScribe v2 Realtime

The base URL can point at an ElevenLabs Data Residency endpoint

Runs the call
Visual workflow builderTransfer Call to a personQA scored per step

Transfer works on Twilio, Telnyx, Plivo, Cloudonix and Asterisk ARI

Runs on
Your own servers with DockerDograh CloudYour carrier's numbers

Self-hosting puts updates and uptime on your team, while Dograh Cloud runs it for you

03 / Why move the agent

Move the agent, keep the voice

The platform's code is yours

Dograh is open source under BSD-2-Clause, so your team can read and change every part of it.

Any model per agent

ElevenLabs can voice one agent while another uses Dograh's models or open-weight models on your servers.

One voice across clips and speech

Pre-recorded clips play where they fit and the cloned voice speaks the rest, so callers hear one voice.

No per-minute platform fee self-hosted

Servers, model keys and phone minutes still cost. Dograh Cloud charges $0.01 a minute for the platform, plus models.

04 / Dograh vs ElevenAgents

What changes when you own the platform

What you getDograhElevenAgents
Open-source code for the whole platform✓✕
Self-host with Docker, no sales contract✓✕
No per-minute platform fee when self-hosted✓✕
05 / Guides

ElevenLabs alternatives, step by step

Questions
Can I use a cloned ElevenLabs voice in Dograh?+

Yes. Paste the voice ID by hand on the voice tab of a BYOK configuration. Voices in your account also load from your key.

Which ElevenLabs models can I pick?+

Voices run from Flash v2.5, the default, up to Eleven v4 and v4 Turbo. Scribe v2 Realtime handles speech-to-text.

Does call audio still go to ElevenLabs?+

Yes. The agent's replies go to ElevenLabs to become speech, and with Scribe the caller's audio goes there too.

Who bills the ElevenLabs usage?+

ElevenLabs bills your own account at its API rates. Self-hosted Dograh adds no platform fee, and Dograh Cloud charges $0.01 a minute.

Can the agent transfer calls to a person?+

Yes, on Twilio, Telnyx, Plivo, Cloudonix and Asterisk ARI. Dograh calls the destination first and connects the caller only when someone answers.

Can Dograh take keypad input?+

No. Dograh has no keypad (DTMF) tool, so callers answer by voice. Build those steps around a spoken check.

Bring your voice ID and make a test call.

Self-host in one command, or start on Dograh Cloud.

$ curl … start_docker.sh && ./start_docker.sh