Nox Voice 1.2 (Beta)

Voice model · Page last updated: September 29, 2026

Talk to Nox in real time

Nox Voice 1.2 (beta) is the family's real-time speech model, built on leading conversational voice AI technology. It runs a live, low-latency speech loop: spoken input in, natural spoken answers out, with the full exchange transcribed into the chat thread as it happens.

Voice 1.2 is a real-time conversation loop rather than text-to-speech bolted onto a chat model. Latency is low enough that the exchange flows at spoken-conversation pace, and the voice output is natural in cadence rather than read-aloud flat.

It is multilingual across 12 spoken languages — English, Spanish, French, German, Italian, Portuguese, Chinese, Japanese, Korean, Hindi, Arabic, and Russian — speaking and listening in whichever one is selected.

It supports interruption: the model listens while it speaks, so talking over it redirects the answer mid-stream instead of waiting for the turn to finish. Every turn on both sides is transcribed live and saved into the chat thread, so the spoken conversation has a complete written record that can be reread later.

Voice sessions run behind the same Leo safety layer as text conversations — the deterministic red-flag detector, with its published recall and false-positive figures, screens voice turns too.

Nox Voice 1.2 is available on MAX and is currently in beta.

Capabilities

• Real-time spoken conversation with natural voice output
• 12 spoken languages, selectable per conversation
• Interruption support — listens while speaking, redirects mid-answer
• Live transcript of every turn, saved into the chat thread

Model details

• Loop — Real-time speech in, real-time speech out — a live conversation, not read-aloud text
• Languages — 12 spoken languages — English, Spanish, French, German, Italian, Portuguese, Chinese, Japanese, Korean, Hindi, Arabic, Russian
• Live transcript — Yes — every turn is transcribed and saved to the chat
• Interruptions — Listens while speaking — talking over Nox redirects the answer mid-stream
• Safety layer — Runs behind Leo, Nox's deterministic red-flag detector, on every turn
• Available on — MAX

Availability: Available on MAX. Currently in beta.

Nox Voice 1.2 — frequently asked questions

What is Nox Voice 1.2?

Nox Voice 1.2 is Nox's voice model. Nox Voice 1.2 (beta) is the family's real-time speech model, built on leading conversational voice AI technology. It runs a live, low-latency speech loop: spoken input in, natural spoken answers out, with the full exchange transcribed into the chat thread as it happens.

What can Nox Voice 1.2 do?

Nox Voice 1.2's capabilities: Real-time spoken conversation with natural voice output; 12 spoken languages, selectable per conversation; Interruption support — listens while speaking, redirects mid-answer; Live transcript of every turn, saved into the chat thread.

Which Nox plans include Nox Voice 1.2?

Available on MAX. Currently in beta.

Related Nox models

• Nox Orion 2 — The upgrade to 1.4 Orion — faster, more accurate, lower cost
• Nox Orion Flash 2 — The fast Orion — full thinking depth, 300B model
• Nox 1.3 Astra — Deep reasoning with a one-million-token context
• Nox Vela 2 — The fast model on every plan — now more accurate
• Nox Voice 1 Advance — Officially launched spoken Nox — live on every plan

Back to the Nox model family · Plans and pricing · Nox features