Mic button
Speak straight into the composer.
Tap the mic and just4o.chat starts speech-to-text inside the chat box itself, so voice input feels like a faster way to write, not a detour into a separate tool.
Voice & TTS
A full voice mode, with the model you chose.
Tap the mic to speak into the composer. Tap the speaker on any answer to hear it back. Turn on voice mode and just4o.chat can auto-send the transcript after your silence window, read replies aloud when they land, and keep the whole exchange inside the same chat, with your selected text model still doing the thinking.
Mic button for speech-to-text in the composerSpeaker button on any assistant replyVoice mode keeps your selected text model activeGenerated audio is saved in Files > Generated Audio
Voice inside the chat surface
Speak.Listen.Continue.
The mic, the speaker, and full voice mode all live inside chat, so voice feels like a first-class way to use just4o.chat instead of a side destination.
Speech in. Reply out. Same thread.
Audio-native in chatGPT audio mini
Choose it as the active chat model for the closest full OpenAI-style voice loop.
OpenAI read-aloudGPT-4o mini TTS
Playback model used from Voice settings.
Grok read-aloudGrok Voice
xAI playback path available in Voice settings.
ElevenLabs read-aloudElevenLabs Flash v2.5
Low-latency ElevenLabs playback path.
How it works in product
Three ways to use voice, depending on what you need right now.
just4o.chat does not force every voice interaction into one rigid mode. You can use the mic, the speaker, or the full voice loop depending on whether you want faster input, easier playback, or a hands-free conversation.
Speaker button
Read any answer back out loud.
Assistant replies keep a speaker action right on the message, which makes it easy to listen to the exact output you just got without changing modes or leaving the thread.
Voice mode
Turn on the full loop when you want it.
With a silence window configured, transcripts can auto-send, replies read aloud when they finish, and the whole exchange stays inside the same chat.
Available voice models
The exact voice stack we ship today.
Your selected text model still stays active in voice mode. These are the actual voice-specific models currently available in just4o.chat for audio-native chat or read-aloud playback.
Audio-native chat
OpenAI chat audio models
Use one of these as the active chat model when you want the closest full OpenAI voice-chat feel inside just4o.chat.
GPT audio minigpt-audio-mini
GPT audio mini · Oct. 2025gpt-audio-mini-2025-10-06
GPT audio mini · Dec. 2025gpt-audio-mini-2025-12-15
OpenAI read-aloud
OpenAI playback models
Selectable in Voice settings for reply playback and previews.
GPT-4o mini TTS (2025-12-15)gpt-4o-mini-tts-2025-12-15
GPT-4o mini TTS (2025-03-20)gpt-4o-mini-tts-2025-03-20
GPT-4o mini TTSgpt-4o-mini-tts
Grok read-aloud
xAI playback model
Selectable in Voice settings for Grok read-aloud.
Grok Voicegrok-tts
11
ElevenLabs read-aloud
ElevenLabs playback model
Selectable in Voice settings for ElevenLabs read-aloud.
ElevenLabs Flash v2.5eleven_flash_v2_5
Voice settings
Configure the voice layer once, then let it disappear.
Voice settings live in Account, where the behavior is explicit. Choose the read-aloud model, pick provider voices, adjust the silence window, switch the thinking sound on or off, and set playback speed without muddying what model is actually answering you.
Text modelYour selected text model stays selected
Voice mode wraps around the active chat model instead of silently swapping it out.
Read-aloud modelOpenAI, Grok Voice, or ElevenLabs playback
Playback is configurable on purpose, so the speaking layer can be different from the thinking layer.
Voice settingsProvider voices, OpenAI styles, silence timing, thinking sound, playback speed
Those controls live in Voice settings in Account, where the voice layer is explicit and adjustable.
OpenAI voice styles
NeutralWarmExpressiveCalmStoryteller
Prefer the standard OpenAI feel?
Set read-aloud to GPT-4o mini TTS for the familiar OpenAI playback voice. If you want the full audio-native OpenAI loop instead of just OpenAI playback, switch the active chat model itself to GPT audio mini.
Saved in Files
Generated audio stays with the rest of the work.
Spoken output is not treated like a disposable playback event. Generated audio is stored in just4o Files under Generated Audio, where it can be found again, replayed, and managed alongside the rest of your generated media.
Files / Generated Audio
Voice-mode and read-aloud output lands in the generated-audio area of just4o Files instead of vanishing after playback.
Saved in Files under Generated Audio
Kept alongside the rest of your generated media
Built to be found and replayed later, not treated like disposable playback
Keep the conversation audible without giving up control.
Open chat, speak into the composer, turn on voice mode when you want the full loop, and keep the reply layer attached to the same model, thread, and files you already use.