On 16 July 2026, xAI announced Grok 4.5. xAI paired high coding-agent scores with a claim of much shorter outputs and lower per-token pricing than premium comparisons.
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 | 16 July 2026 |
| Availability or weight release | 16 July 2026 through xAI services and API channels |
| Release type | proprietary general-purpose coding and agent model |
| Access | hosted xAI products and API; no open weights |
| Architecture | an undisclosed Grok frontier model tuned for coding agents and efficient reasoning |
| Maximum stated context | the Grok 4.5 limit documented by xAI |
What changed
Grok 4.5 makes concision part of the capability claim. xAI reports an average of 15,954 output tokens on its comparison against 67,020 for Claude Opus 4.8, about 4.2 times fewer, while remaining competitive on terminal and SWE tasks. That can lower cost and latency, but only if shorter trajectories do not hide missed verification or brittle patches.
The practical comparison is therefore not simply whether Grok 4.5 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 |
|---|---|---|
| Terminal-Bench 2.1 | 83.3 | Command-line agent result reported by xAI |
| SWE-Bench Pro | 64.7 | Repository-level software engineering result |
| DeepSWE 1.0 | 62 | Long-horizon coding result; version must be preserved |
These are release-time results, not independently reproduced guarantees. xAI’s benchmark table uses named agent harnesses and vendor-reported comparison values; token counts should be compared with identical task completion criteria. 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 4.5 is described as an undisclosed Grok frontier model tuned for coding agents and efficient reasoning with the Grok 4.5 limit documented by xAI of stated context. Its access position at verification time is hosted xAI products and API; no open 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 4.5, 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
Grok 4.5 is a serious proprietary coding-agent contender. The right test is completed, reviewed work per dollar, with explicit checks for whether its shorter outputs omit evidence or validation.
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.
