Your summary is ready. Want me to send it to the team?
PuckVoice / Alluqi Bridge
A voice of your own.
A conversation. On your terms.
Talk to Bridge naturally, with speech running on your own hardware.
- Speech stays on your own hardware
- No per-minute fees
- Interrupt anytime
- Tools work mid-conversation
- Choose your voice
Bridge / Voice
Listening
Scripted example. You ask for a project plan. Bridge starts a reply. You interrupt to ask for a shorter plan, and Bridge follows your new direction.
Talk to Bridge
Made to be heard
Find your voice.
Hear a few of Bridge’s local voices. Pick a voice that feels right for your conversation.
I found three open slots on Thursday afternoon.
HeartDone. The report is saved in your project folder.
Aiden01 / Voice in Bridge
Start voice in the chat you already have open. Keep the context, and keep the transcript as normal messages.
- Start in your current conversation
- Move between speaking and typing
- Return to the transcript later
Bridge / Voice
Listening
Scripted example. You ask for a project plan. Bridge starts a reply. You interrupt to ask for a shorter plan, and Bridge follows your new direction.
Talk to Bridge
02 / Voice in Bridge
Change direction while Bridge is speaking. Add a detail, correct the request or ask a follow-up without starting over.
- Speak to interrupt a reply
- Keep the thread of the conversation
- Continue with your new direction
Bridge / Voice
Listening
Scripted example. You ask for a project plan. Bridge starts a reply. You interrupt to ask for a shorter plan, and Bridge follows your new direction.
Talk to Bridge
03 / Voice in Bridge
Ask Bridge to use your connected tools during the conversation. Review actions that need your approval before they proceed.
- Use tools in the current chat
- Keep approvals in the conversation
- Review the result in your transcript
Bridge / Voice
Listening
Scripted example. You ask for a project plan. Bridge starts a reply. You interrupt to ask for a shorter plan, and Bridge follows your new direction.
Tool requestSave a project noteApproval required before writing
Talk to Bridge
04 / Voice in Bridge
Pick from the voices available in your local speech setup. Find a voice you want to hear throughout the day.
- Choose a local voice
- Use your configured speech models
- Change your selection in settings
Bridge / Voice
Listening
Scripted example. You ask for a project plan. Bridge starts a reply. You interrupt to ask for a shorter plan, and Bridge follows your new direction.
Local voice
Talk to Bridge
05 / Voice in Bridge
Use the composer microphone when you want to speak a message instead of typing. Review the text before you send it.
- Dictate into the chat composer
- Review and edit your words
- Speak from the pet, too
Bridge / Voice
Listening
Scripted example. You ask for a project plan. Bridge starts a reply. You interrupt to ask for a shorter plan, and Bridge follows your new direction.
Chat composer · dictation
Draft a short note for the project.
Review your words before sending.Talk to Bridge
06 / Voice in Bridge
Open the Mini mic from your Bridge pet. Keep a small entry point to voice within reach while you work.
- Start from the pet
- Use the Mini microphone
- Continue in your Bridge conversation
Bridge / Voice
Listening
Scripted example. You ask for a project plan. Bridge starts a reply. You interrupt to ask for a shorter plan, and Bridge follows your new direction.
Bridge petMini mic
Talk to Bridge
07 / Voice in Bridge
Speech-to-text and text-to-speech run on your server. Your local speech path sends no audio to a cloud service.
- Transcribe on your own hardware
- Generate speech on your server
- Choose which tools and models you connect
Bridge / Voice
Listening
Scripted example. You ask for a project plan. Bridge starts a reply. You interrupt to ask for a shorter plan, and Bridge follows your new direction.
Your microphoneYour server · speech modelsYour Bridge conversation
Talk to Bridge
Your hardware. Your conversation.
Pay once. Keep talking.
Pay for the hardware once. No per-minute fees for local speech.
Alluqi Bridge
Pay for the hardware once.
No per-minute fees.
Hardware, electricity, setup and maintenance still have costs. Model licenses apply. External models and tools may charge separately.
Hosted voice APIs
Selected USD list prices. Billing units and services differ; these are not equivalent complete-conversation costs or a quality comparison.
x.ai / Grok Voice 3.0 · speech to speech$4.80 / hour. Text input: $0.004 per input; tool costs may apply separately.
$0.08 / minute · Official pricing
x.ai / Speech to textTranscription only; not a complete voice conversation.
$0.10 / hour REST · $0.20 / hour streaming · Official pricing
x.ai / Text to speechSpeech generation only; not a complete voice conversation.
$15 / million characters · Official pricing
OpenAI / GPT-Live 1 · voice sessionsBilled per second. Backend model and tool usage are charged separately.
$0.05 / minute · Official pricing
Prices as of October 26, 2026. Check the linked source for current rates and additional charges.
A few details
Common questions.
About speaking with Bridge.
Which models does voice use?
Bridge uses the speech-to-text and text-to-speech models available in your local setup, alongside your chat model. Check your server configuration for the models and voices you can use.
What hardware do I need?
You need Bridge, a microphone and a server running your speech and chat models. Hardware requirements depend on those models and what else the machine is running. Contact us to plan a setup for your workflow.
Which languages are supported?
Language support depends on your selected speech models. Choose a language supported by both your transcription and speech setup, and test it with the material you use.
Can I choose the voice?
Yes. Choose from the voices available in your configured local speech setup. Available voices depend on the model and server you use.
Where does my speech go?
In the local speech configuration, transcription and speech generation run on your server. Transcripts become normal chat messages. External chat models and connected tools can send relevant information to their configured endpoints; review those connections separately.
What does the conversation feel like?
Voice is designed for natural back-and-forth: speak, listen and interrupt to change direction. Responsiveness depends on your hardware, models and network.
How is dictation different from voice mode?
Dictation turns speech into text you can review in the composer. Voice mode carries on a spoken conversation and keeps its transcript in the chat.
Can I use it commercially?
Commercial use depends on the licenses of the speech and chat models you choose, and the rights to the material you use. Check those terms before using a setup for business.
Let’s talk.
Your voice. Your hardware. Your Bridge.