Microsoft Azure OpenAI Makes GPT-6 Sol & Luna Generally Available in Foundry

Microsoft has expanded its GPT-6 series with the general availability of GPT-6 Sol and GPT-6 Luna in Microsoft Foundry. These models complement GPT-6 Astra, offering optimized capabilities for production AI agents that reduce noise and improve task completion.

GPT-6 Sol is designed for enterprise agents, coding, and complex knowledge work, including multi-step reasoning, long-context analysis, and tool-based workflows. It serves as a cost-effective, frontier-efficient option for production workloads or migrations from legacy models.

GPT-6 Luna, positioned as Sol’s smaller and faster sibling, is built for high-volume tasks such as data extraction, summarization, request routing, and routine customer interactions. It enables efficient intelligence for preparatory work where deeper reasoning is not always required.

GPT-6 Astra, Sol, and Luna are available with Standard deployment across 28 Global regions, US, and EU Data Zones. Provisioned Throughput is available for Astra and Sol, and Priority Processing for Sol, offering tailored capacity and latency for critical production demands.

GPT-6 models in Azure OpenAI include built-in safety training, content filters, and guardrails for prompts and outputs. Foundry further applies controls at key checkpoints, identity/access governance, and Microsoft Purview policies, supporting continuous safety strengthening.