On 29 July 2026, xAI announced Grok Voice Think Fast 2.0. The model improved conversational dynamics, tool use and multilingual transcription while reducing latency and reasoning-token use.
This release brief was checked against first-party material on 10 August 2026. The date above is the public announcement date, not the date a repository was created or a third-party provider added the model. Where access or weights arrived later, that distinction is recorded below.
Release record
| Field | Verified detail |
|---|---|
| Announcement | 29 July 2026 |
| Availability or weight release | 29 July 2026 by API, with the latest alias scheduled to switch on 5 August |
| Release type | proprietary speech-to-speech voice-agent model |
| Access | xAI Voice API; no released weights |
| Architecture | full-duplex speech model that performs reasoning in parallel with spoken output |
| Maximum stated context | session limit not stated as a headline figure in the release article |
What changed
xAI published a detailed comparison across overall speech quality, reasoning, full-duplex conversation and voice-agent task completion. It also documented the model-alias migration date and fixed per-minute pricing, making this a distinct operational model release rather than a silent voice update.
The practical comparison is therefore not simply whether Grok Voice Think Fast 2.0 has the largest headline score. Teams need to ask whether its architecture, access terms, latency, tool behaviour and evaluation setup match the workload they actually intend to run. A model can lead one harness while losing on cost, refusal behaviour, multilingual quality or repeatability in another.
Benchmarks worth retaining
| Evaluation | Reported result | How to read it |
|---|---|---|
| AA Speech-to-Speech Quality Index | 82.9% | Compared with 75.7% for Think Fast 1.0 |
| Tau-voice Bench | 56.5% | Agentic voice performance reported by Artificial Analysis |
| Time to first audio | 0.70 seconds | Compared with 1.25 seconds for the previous xAI model |
These are release-time results, not independently reproduced guarantees. The launch mixes Artificial Analysis scores with xAI multilingual transcription tests and production A/B results, so each number must retain its original harness. Scores should remain attached to the disclosed effort setting, agent harness, tool access, timeout, context-management policy and judge model. Moving a number into a procurement sheet without those conditions creates false comparability.
Architecture and access
Grok Voice Think Fast 2.0 is described as full-duplex speech model that performs reasoning in parallel with spoken output with session limit not stated as a headline figure in the release article of stated context. Its access position at verification time is xAI Voice API; no released weights. That wording matters: open weights, source-available weights, an API, a product preview and a research demonstration give adopters very different rights and different levels of reproducibility.
Before deployment, record the exact model identifier or checkpoint, inference stack, quantisation, reasoning setting, region, price schedule and supplier terms. If the release uses a custom licence, read the licence itself rather than relying on the word “open” in launch copy. If it is API-only, preserve the dated documentation and change-notice route because the served snapshot can change without a downloadable artefact.
What an evaluation should test next
For Grok Voice Think Fast 2.0, a credible internal gate should include:
- a frozen set of representative tasks with pass, fail and abstain criteria;
- a matched baseline using the same tools, timeout, prompt budget and reviewer rubric;
- repeated runs to expose variance rather than reporting a single best attempt;
- latency, token use and total task cost alongside task success;
- adversarial, multilingual and long-context cases relevant to the real deployment; and
- rollback evidence showing the previous model can be restored safely.
The wider model change-control guide explains how to keep model, prompt, tool and corpus changes reconstructable. The AI dependency inventory guide covers the release and supplier records needed after deployment.
AIEngine verdict
Voice Think Fast 2.0 deserves a separate brief. Teams should pin the checkpoint, because the latest alias changed automatically, and reproduce both noisy transcription and tool-call timing.
This is a launch assessment, not a certification. Benchmark leadership is useful evidence of where to test; it is not authorization to place the model in a high-impact workflow without domain evaluation, security review and an accountable owner.
Primary sources
Image provenance
Hero image: xAI official release artwork. The locally served WebP is a crop of the first-party release or model-card asset recorded in the repository provenance manifest.
