Grok Voice 3.0
Run full-duplex speech-to-speech voice agents with Grok Voice 3.0 — sub-second replies, top-ranked call accuracy, and API access from $0.08 per minute.
Support
Pro AI Tools
Explore elite tools including the Suno AI Music Generator.
MiniMax H3
MiniMax H3 AI Video Generator
Seedance 2.5
The Future of AI Video Is Here.
Seedance 2.0
The Future of AI Video Is Here.
Veo3.1
Create Stunning Videos with Veo3.1
FLUX 3 Video Generator
Kling 3.0
Next-Gen AI Video Generator
Grok Video Generator
Create Videos from Text or Images with AI
MiniMax H3 video generator
MiniMax H3 AI Video Generator
A Closer Look at Grok Voice 3.0
Built for live conversation, Grok Voice 3.0 pairs benchmark-leading recognition accuracy with near-instant replies and production-grade reliability at enterprise scale.
- Benchmark-Leading Call AccuracyIndependent evaluations place Grok Voice 3.0 first on the Speech-to-Speech Index at 86.9, a composite of speech reasoning, τ-Voice, and blind head-to-head preference.
- Conversations That Feel InstantReplies land in well under a second, and a 97% Full Duplex Bench score means the agent handles pauses, turn shifts, and interruptions the way a person would.
- Battle-Tested at Global ScaleBehind millions of daily exchanges across Tesla and SpaceX, including 20,000 Starlink support calls a day — roughly 70% closed with no human stepping in.
How to Deploy Grok Voice 3.0
From your first API call to multilingual call automation, here is the short path to shipping a production voice agent with this model.
Grok Voice 3.0 Capabilities That Matter in Production
Accuracy, response speed, noise tolerance, and cost per task decide whether a voice agent survives real traffic. Here is how this model measures up.
Record-Setting Call Resolution
Grok Voice 3.0 posts the best customer-support resolution rate measured to date and holds first place on the Speech-to-Speech Index at 86.9.
Human-Like Turn Taking
A 97% Full Duplex Bench result shows how gracefully the agent yields the floor, waits out pauses, and recovers when a caller cuts in.
Holds Up Under Messy Audio
The τ-Voice suite throws varied accents, noisy backgrounds, and interruptions at the agent; Grok Voice 3.0 comes out at 71.6%.
Smart Without the Lag
On τ-Voice it keeps pace with GPT-Live-1 (Astra) while answering about 40% faster, widening what an agent can be both smart and quick enough to do.
Leaner Spend per Task
In the Big Bench Audio run from Artificial Analysis, its per-task spend lands below both GPT-Live-1 (Astra, Medium) and GPT-Realtime-2.1.
Strong Enterprise Accuracy
On ServiceNow's EVA-Bench it reaches 74% — the highest accuracy recorded, and the best combined accuracy plus customer-experience result.
Grok Voice 3.0 Questions, Answered
Quick answers on accuracy scores, response times, per-minute pricing, and how this model performs on live customer calls.
What exactly is Grok Voice 3.0?
It is a speech-to-speech model purpose-built for live voice agents, released in the API at the same per-minute rate as the previous generation.
How accurate is it on independent benchmarks?
It sits at the top of Artificial Analysis's Speech-to-Speech Index with 86.9 and records 97% on Full Duplex Bench.
What latency can I expect during a live call?
Replies arrive in under one second, and on τ-Voice it responds roughly 40% faster than GPT-Live-1 (Astra).
What does it cost to run?
The price sheet did not move: the Voice Agent API bills $0.08 per minute ($4.80 per hour), the same figure Grok Voice 2.0 charged.
Does it work for real customer support?
It does — Starlink routes 20,000 customer calls a day through it in multiple languages, and around seven in ten close without a human.
Will my existing grok-voice-latest agent break?
No. Agents already pointing at grok-voice-latest were switched over behind the scenes, and nothing in your integration needs to change.
Put Grok Voice 3.0 to Work
Ship a voice agent that answers in under a second, resolves calls at enterprise scale, and still bills at $0.08 per minute through the Voice Agent API.