Microsoft Brings xAI’s Grok 4 Fast Models to Azure AI Foundry
Readers help support Windows Report. We may get a commission if you buy through our links.
Read our disclosure page to find out how can you help Windows Report sustain the editorial team. Read more
Microsoft has announced that the Grok 4 Fast models from xAI are now available in preview on Azure AI Foundry. Here, we’re talking about two models, grok-4-fast-reasoning and grok-4-fast-non-reasoning. When xAI launched these models, it said that both offer rapid AI performance, multimodal input, and parallel task execution.
The Grok 4 Fast models are personalized for scenarios where speed and orchestration are two of the most important aspects. Both options support structured outputs, JSON schema integration, long-context processing up to 131K tokens, and H100 GPU efficiency.
The reasoning model is further optimized for agentic decision-making, while the non-reasoning variant offers a streamlined option for specific workloads. Microsoft says that running Grok 4 Fast on Azure also provides:
- Global availability and reliability through its cloud footprint
- Enterprise-grade security, identity, and encryption
- Centralized governance and cost controls via the Azure Portal
- Integration with services like Synapse, Cosmos DB, and Logic Apps
- 24/7 enterprise support and SLAs
As far as use cases are concerned, developers can use the models for:
- Real-time orchestration
- Multimodal document analysis
- Enterprise search
- Parallel tool integration
- Conversational AI and decision support.
Speaking of pricing, grok-4-fast-reasoning starts at $0.43 per 1M input tokens and $1.73 per 1M output tokens. The Grok 4 Fast models are live in the Azure AI Foundry Model Catalog, where developers can deploy them immediately. Just in case you’re looking to learn more about GrokCode Fast 1, here’s our explainer guide for you.
