SpaceXAI has released its latest real-time voice model, Grok Voice 3.0. This speech-to-speech model is designed to listen, reason, call tools, and speak. It is available via SpaceXAI’s hosted Voice Agent API.
Key points
- Grok Voice 3.0 ranks first on the Artificial Analysis Speech-to-Speech Index.
- The model achieved a score of 86.9 on the AA Speech-to-Speech Index.
- Grok Voice 3.0 is a speech-to-speech model that listens, reasons, and speaks.
- It did not achieve first place in all individual benchmark components.
Grok Voice 3.0 has achieved the highest score on the Artificial Analysis Speech-to-Speech Index, with a score of 86.9, ranking it first among currently evaluated models. This index is an equal-weighted average of four components: Big Bench Audio reasoning, τ-Voice agentic performance, Speech Agent Arena preference, and task success rate.
While Grok Voice 3.0 achieved a top composite score, it did not lead in all individual benchmarks. It scored 98.5% on Big Bench Audio, ranking fourth, and 97.0% on the Full Duplex Bench, also ranking fourth. Its task success rate was 94.8%.
By the numbers
- AA Speech-to-Speech Index score
- 86.9
- time to first audio
- 0.81 s
- task success rate
- 94.8%
The model's reasoning capabilities are enabled by default with a "high" effort setting, which can be adjusted to "none" for quicker responses. Server-side voice activity detection supports barge-in functionality. Grok Voice 3.0 can integrate with various tools, including collections search, web search, X search, remote MCP servers, and custom functions.
Sources and references
Original report
Meet Grok Voice 3.0: SpaceXAI Voice Agent Model Ranks #1 on Artificial Analysis
MarkTechPost
Read on MarkTechPost (opens in a new tab)Linked in the original report
- Linkx.ai/news/grok-voice-3 (opens in a new tab)
- Linkartificialanalysis.ai/speech-to-speech (opens in a new tab)
- Linkdocs.x.ai/developers/model-capabilities/audio/spe… (opens in a new tab)
- Linkservicenow.github.io/eva (opens in a new tab)
- Linkdevelopers.openai.com/api/docs/models/gpt-live-1 (opens in a new tab)
- Linkai.google.dev/gemini-api/docs/pricing (opens in a new tab)
Ref 001N4NMU
Cite this
APA
Interestana AI Editorial. (2026, October 15). SpaceXAI's Grok Voice 3.0 Ranks First on Speech Index. Interestana. https://interestana.com/articles/meet-grok-voice-30-spacexai-voice-agent-model-ranks-1-on-artificial-analysis-001n4nmu
MLA
“SpaceXAI's Grok Voice 3.0 Ranks First on Speech Index.” Interestana, 15 Oct. 2026, interestana.com/articles/meet-grok-voice-30-spacexai-voice-agent-model-ranks-1-on-artificial-analysis-001n4nmu.
Chicago
Interestana AI Editorial. “SpaceXAI's Grok Voice 3.0 Ranks First on Speech Index.” Interestana, October 15, 2026. https://interestana.com/articles/meet-grok-voice-30-spacexai-voice-agent-model-ranks-1-on-artificial-analysis-001n4nmu.
BibTeX
@misc{interestana_001n4nmu,
author = {{Interestana AI Editorial}},
title = {{SpaceXAI's Grok Voice 3.0 Ranks First on Speech Index}},
howpublished = {Interestana},
year = {2026},
month = oct,
url = {https://interestana.com/articles/meet-grok-voice-30-spacexai-voice-agent-model-ranks-1-on-artificial-analysis-001n4nmu},
note = {Summary of reporting by MarkTechPost, https://www.marktechpost.com/2026/10/14/meet-grok-voice-3-0-spacexai-voice-agent-model-ranks-1-on-artificial-analysis/}
}Credit the source
Original reporting: MarkTechPost, “Meet Grok Voice 3.0: SpaceXAI Voice Agent Model Ranks #1 on Artificial Analysis”, https://www.marktechpost.com/2026/10/14/meet-grok-voice-3-0-spacexai-voice-agent-model-ranks-1-on-artificial-analysis/(opens in a new tab)
Follow this story
Get Artificial Intelligence in your inbox — new stories in 40 words or fewer, once a week.
Subscribe