Apresentando o Grok 4.7
Nosso modelo mais capaz para programação de longa duração e trabalho intelectual.
Grok 4.7
A frontier model for long-running coding and knowledge work, at the same price and speed as Grok 4.6.
Announcements
Model card do Grok 4.7
O model card oficial do Grok 4.7, com os benchmarks de capacidade do modelo e as avaliações de salvaguardas.
Cursor firma parceria com a SpaceX para treinamento de modelos
A Cursor firmou parceria com a SpaceX para acelerar nossos esforços de treinamento de modelos.
Built for coding and knowledge work alike
Long-horizon problemsLong-horizon
problems
Grok 4.7 stays with difficult tasks across many steps. It uses tools, checks its own work, adjusts its approach, and keeps moving toward a finished result.
Ambitious engineeringAmbitious
engineering
Grok 4.7 works across large codebases and extended engineering projects. It researches unfamiliar systems, edits across files, runs tests, and verifies the result.
In-depth knowledge workIn-depth
knowledge work
Beyond code, Grok 4.7 works across documents, spreadsheets, presentations, and other professional artifacts. Give it a multi-step project that requires research, analysis, and synthesis.
Checks its own workChecks its
own work
Grok 4.7 verifies results more carefully than Grok 4.6 and manages longer context, with a 256k standard window and 500k long context.
Built for more than software engineering
Grok 4.7 is built for long-running coding and knowledge work. It uses a larger base than Grok 4.6 and a longer training run on harder, multi-hour tasks. It handles projects that span research, analysis, implementation, and several rounds of refinement.
On CursorBench 4.0 it scores 46.3% at extra high effort. It is served at the same price and speed as Grok 4.6, and Cursor subscription plans for individuals and teams include significant usage of the model.
A strong foundation
We trained Grok 4.7 jointly with SpaceXAI on a new, larger base model than Grok 4.6. The reinforcement learning run was longer and weighted toward problems that take many hours to complete.
The model is better at verifying its own work and managing longer context. It was also trained to understand the Grok Bot harness, which improves conversational tasks and general knowledge work.
Benchmarks
| Grok 4.7 xHigh | Grok 4.6 High | GPT-5.6 Sol Max | Fable 5.1 Max | |
|---|---|---|---|---|
| CursorBench 4.0Software engineering | 46.3% | 40.4% | 41.7% | 51.8% |
| DeepSWE v1.1Software engineering | 71.0% high effort | 65.2% | 72.7% | 70.0% |
| EEBenchElectrical engineering | 64.0% | 53.0% | 39.4% | 56.4% |
| AA Briefcase v1.1Multi-hour office work | 1,657 | 1,546 | 1,487 | 1,678 |
| Terminal-Bench 4.0Multi-hour terminal work | 37.6% | 20.3% | 37.3% | 57.9% |
| Harvey Legal Agent BenchmarkLegal work | 19.6% | 15.8% | 2.5% | 6.7% |
| HealthBench ProfessionalClinical reasoning | 56.7% | 48.5% | 60.5% | 62.1% |
Resultados divulgados no anúncio do Grok 4.7. O resultado do Grok 4.7 no DeepSWE foi obtido com esforço alto.
No GDPval, o Grok 4.7 atinge 1.695 pontos Elo com esforço extra-alto, contra 1.605 do Grok 4.6, 1.735 do Fable 5.1 e 1.542 do GPT-6 Astra.
(Above) Grok 4.7 results across agentic coding and knowledge work benchmarks.
Available everywhere you work
Grok 4.7 is available today in Cursor across desktop, web, iOS, CLI, and our SDK.
Interactive demo with multiple windows showing Cursor's AI-powered features. The interface is displayed over a subtle, solid brand background.
Set up Datadog APM for all API routes
Explored 12 files, 3 searches
I'll add the Datadog SDK, instrument all API routes with tracing, and set up error tracking.
Worked for 7m 42s
Processed screen recording
Done — here's the Datadog dashboard showing all instrumented routes.
Acme Labs
Summary
Integrated Datadog APM across all 24 API routes with error tracking and custom metrics.
Add a follow up...
This element contains an interactive demo for sighted users. It's a demonstration of Cursor integrated within Slack, showing AI-powered assistance inside team communication. The interface is displayed over a subtle, solid brand background.
FAQ
Grok 4.7 is a frontier model from Cursor and SpaceXAI for coding and knowledge work. It builds on Grok 4.6 with a larger base, longer training on harder multi-hour tasks, and stronger self-verification.
Grok 4.7 is available today in Cursor across desktop, web, iOS, CLI, and our SDK. You can also build your own agents on top of Grok 4.7 with our SDK docs.
Grok 4.7 is part of the Cursor Models pool on individual and team plans, alongside Grok 4.6, Grok 4.5, and Composer 2.5. It is priced the same as Grok 4.6. Standard on-demand usage is $2/M input, $0.50/M cached input, and $6/M output tokens. The Fast variant is $4/M input, $1/M cached input, and $12/M output tokens. See the model docs for full details.