Today, Grok 4.7 from SpaceXAI is generally available in Microsoft Foundry Models, expanding the portfolio of leading models available to developers and enterprises building the next generation of AI applications. Based on SpaceXAI's latest model release, Grok 4.7 is designed for advanced reasoning, long-horizon problem solving, software engineering, and agentic workflows, making it particularly well suited for complex technical and knowledge-intensive tasks.
As organizations move beyond chat experiences toward autonomous agents and AI-powered workflows, demand continues to grow for models that can reason through multi-step problems, effectively use tools, and maintain context across longer tasks. Grok 4.7 was developed with these scenarios in mind, incorporating enhanced training focused on reasoning, software engineering, technical domains, productivity workflows, and agentic reinforcement learning environments.
Built for advanced engineering workflows
Grok 4.7 was designed to tackle complex technical tasks that require extended reasoning, planning, and execution. In addition to strong coding capabilities, the model was trained and evaluated across software engineering, hardware design, Computer-Aided Design (CAD) generation, and other engineering-focused domains
Benchmark highlights
- 71.0% on DeepSWEv1.1
Benchmarks are from model provider (SpaceXAI).
DeepSWEv1.1 evaluates coding agents on 113 original, long-horizon software engineering tasks across 91 active repositories and five languages.
Beyond software, Grok 4.7 demonstrated strong results on CADGenBench, which measures a model's ability to generate executable CAD geometry from design requirements. This reflects SpaceXAI's broader focus on applying AI to engineering and technical problem solving in domains that extend beyond code generation.
Supporting advanced knowledge work
Grok 4.7 also demonstrates strong performance across knowledge-work scenarios that require planning, analysis, and synthesis of information. The model card highlights evaluations covering professional workflows that involve handling documents, structured information, and multi-step reasoning tasks.
These strengths make the model a compelling option for organizations building research assistants, enterprise copilots, analyst tools, and workflow automation solutions. As businesses increasingly look to automate repetitive tasks while helping employees focus on higher-value work, models that combine reasoning, planning, and execution become increasingly valuable.
Benchmark highlights
- 19.6% on Harvey’s final score
Benchmarks are from model provider (SpaceXAI)
The Legal Agent Benchmark evaluates long-horizon legal work organized around realistic client matters. Agents use six file and terminal tools plus document, presentation, and spreadsheet skills to produce work against task-specific criteria.
Pricing
|
Deployment |
Context Window |
Input/1M Tokens |
Output/1M Tokens |
Cache/1M Tokens |
|
Global Standard |
Short context <200K tokens |
$2.00 |
$6.00 |
$0.50 |
|
Long context >=200K tokens |
$4.00 |
$12.00 |
$1.00 | |
|
US DataZone |
Short context <200K tokens |
$2.20 |
$6.60 |
$0.55 |
|
Long context >=200K tokens |
$4.40 |
$13.20 |
$1.10 |
Built for developers
Grok 4.7 is available in Microsoft Foundry through Global Standard and US Data Zone Standard deployments, giving developers flexibility in how they deploy the model for their applications.
Developers can integrate Grok models using familiar Chat Completions and Responses APIs, with support for capabilities such as tool calling and streaming. Foundry also supports authentication using Microsoft Entra ID or API keys, making it easier to integrate Grok into existing applications and agentic workflows.
Grok 4.7 is also planned to support up to 500K context from launch, enabling developers to work across larger codebases, documents, and long-running workflows
Enterprise-ready through Foundry
By bringing Grok 4.7 into Foundry, developers gain access to the model through a unified platform designed for enterprise AI development.
With Foundry, teams can discover, evaluate, deploy, and govern AI models using a consistent development experience. Customers can compare Grok 4.7 with other leading foundation models, integrate it into existing applications and agents, and take advantage of Foundry's tools for observability, evaluation, and responsible AI practices.
This flexibility is particularly important as organizations adopt multi-model strategies, selecting different models for different use cases rather than relying on a single model across all workloads. Foundry makes it easier to experiment, iterate, and operationalize AI solutions while maintaining enterprise-grade governance and control.
Get started
Grok 4.7 is now generally available in Microsoft Foundry Models. Developers can deploy the model, evaluate it against their own scenarios, and integrate it into applications and agents using Foundry APIs.
→ Follow the guide: Deploy and use Grok models in Foundry for integrating Grok into your applications.
→ Compare models: Use Foundry's evaluation tools to benchmark Grok 4.7 against other models on your specific use cases.
Microsoft