Skip to main content
This article lists the retirement schedule for Foundry Models, including the current lifecycle stage, retirement date, and suggested replacement. Use it to plan migrations before a model is deprecated or retired. For details on what each lifecycle stage means and how notifications work, see Microsoft Foundry Models lifecycle and support policy. The Replacement column identifies Microsoft’s recommended replacement and, where supported, the target for automatic upgrades at retirement. You don’t need to wait for a replacement to be listed before evaluating or manually migrating to another compatible model. Evaluate available models using your application and data, comparing quality, latency, and cost. For selection criteria and the full migration process, see Model migration process.

Foundry Models sold by Azure

This section lists the retirement lifecycle for Foundry Models sold by Azure.

Azure OpenAI

Fine-tuned models

Fine-tuned models retire in two phases: training and deployment. Unless explicitly stated, training retires no earlier than the base model retirement date. 1 For existing customers only. Otherwise, training retirement occurs at base model retirement.

Black Forest Labs

Cohere

1 Available through the Hugging Face collection, for managed compute deployment.

DeepSeek

Meta

Microsoft

Mistral AI

MoonshotAI

OpenAI-OSS

xAI

Foundry Models from partners and community

This section lists the retirement lifecycle for Foundry Models sold by partners via Azure Marketplace.

Anthropic

Microsoft Foundry offers Claude models in two versions:
  • Version 1: Hosted on Anthropic infrastructure
  • Version 2: Hosted on Azure

Cohere

Fireworks

Fireworks models on Standard (Per-Token) inference offerings are subject to a 15-day notice period prior to model retirement. Plan your deployments accordingly and monitor notifications for upcoming retirement dates.

Meta

Microsoft

Mistral AI

Nixtla

NTT Data

StabilityAI