Note
Access to this page requires authorization. You can try signing in or changing directories.
Access to this page requires authorization. You can try changing directories.
This article is the reference for the models that model router in Microsoft Foundry can route to, the regions where you can use it, and how long each model router version is supported. Use it to choose models for a model subset, confirm that your region is supported, and plan version moves. Other model router articles link here instead of repeating these lists.
- As of November 18, 2026, model router version
2025-11-18routes across 45 models from OpenAI, Anthropic, Microsoft AI, xAI, DeepSeek, Meta, and Fireworks. Eight of them are new in November 2026. - You don't deploy the underlying models yourself, except Claude models. Deploy each Claude model in the same Foundry resource before model router can route to it.
- Model router routes to the Fireworks models only when Fireworks is enabled for your Azure subscription. See Fireworks models.
- Model router is available in 32 regions for Global Standard deployments, and 23 of them also offer Data Zone Standard. Auto (model router without a deployment) is available in all 32.
- Each model router version has its own retirement date. Model router itself isn't retiring.
Supported models
By default, a model router deployment routes across every model in this table that's available in your region. Auto does the same, unless an admin sets Auto's models for the resource. To route across fewer models, choose a model subset.
When you list models in a deployment request (properties.routing.models) or in Auto's resource settings (properties.instant.routing.models), use the exact format, name, and version values from this table. Per-request controls (routing_config.models) take the name only.
As of November 18, 2026, model router version 2025-11-18 routes across these 45 models:
| Provider | Model (name) |
Version (version) |
Format (format) |
Notes |
|---|---|---|---|---|
| OpenAI | gpt-6.1-terra |
2026-10-26 |
OpenAI |
New in November 2026. |
| OpenAI | gpt-6.1-sol |
2026-09-29 |
OpenAI |
New in November 2026. |
| OpenAI | gpt-6-luna |
2026-09-22 |
OpenAI |
Added September 29, 2026. |
| OpenAI | gpt-6-sol |
2026-09-22 |
OpenAI |
Added September 29, 2026. |
| OpenAI | gpt-6-astra |
2026-09-03 |
OpenAI |
|
| OpenAI | gpt-5.6-luna |
2026-07-09 |
OpenAI |
|
| OpenAI | gpt-5.6-sol |
2026-07-09 |
OpenAI |
|
| OpenAI | gpt-5.6-terra |
2026-07-09 |
OpenAI |
|
| OpenAI | gpt-5.5 |
2026-04-24 |
OpenAI |
|
| OpenAI | gpt-5.4-mini |
2026-03-17 |
OpenAI |
|
| OpenAI | gpt-5.4-nano |
2026-03-17 |
OpenAI |
|
| OpenAI | gpt-5.4 |
2026-03-05 |
OpenAI |
|
| OpenAI | gpt-5.2 |
2025-12-11 |
OpenAI |
|
| OpenAI | gpt-5.1 |
2025-11-13 |
OpenAI |
New in November 2026. |
| OpenAI | gpt-5 |
2025-08-07 |
OpenAI |
|
| OpenAI | gpt-5-mini |
2025-08-07 |
OpenAI |
|
| OpenAI | gpt-5-nano |
2025-08-07 |
OpenAI |
|
| OpenAI | o4-mini |
2025-04-16 |
OpenAI |
Retires November 19, 2026, and leaves the pool then. |
| OpenAI | gpt-4.1 |
2025-04-14 |
OpenAI |
|
| OpenAI | gpt-4.1-mini |
2025-04-14 |
OpenAI |
|
| OpenAI | gpt-4.1-nano |
2025-04-14 |
OpenAI |
|
| OpenAI | gpt-4o |
2024-11-20 |
OpenAI |
|
| OpenAI | gpt-4o-mini |
2024-07-18 |
OpenAI |
|
| OpenAI (open-weight) | gpt-oss-120b |
1 |
OpenAI-OSS |
|
| Anthropic | claude-opus-5-5 |
1 |
Anthropic |
New in November 2026. Deploy it first. See Claude models. |
| Anthropic | claude-sonnet-5-5 |
1 |
Anthropic |
New in November 2026. Deploy it first. See Claude models. |
| Anthropic | claude-haiku-5-5 |
1 |
Anthropic |
New in November 2026. Deploy it first. See Claude models. |
| Anthropic | claude-fable-5-1 |
1 |
Anthropic |
Deploy it first. See Claude models. |
| Anthropic | claude-opus-5 |
1 |
Anthropic |
Deploy it first. See Claude models. |
| Anthropic | claude-sonnet-5 |
1 |
Anthropic |
Deploy it first. See Claude models. |
| Anthropic | claude-opus-4-8 |
1 |
Anthropic |
Deploy it first. See Claude models. |
| Anthropic | claude-opus-4-7 |
1 |
Anthropic |
Deploy it first. See Claude models. |
| Anthropic | claude-opus-4-6 |
1 |
Anthropic |
Deploy it first. See Claude models. |
| Anthropic | claude-sonnet-4-5 |
20250929 |
Anthropic |
Deploy it first. See Claude models. |
| Anthropic | claude-haiku-4-5 |
20251001 |
Anthropic |
Deploy it first. See Claude models. |
| Microsoft AI | mai-code-2-flash |
2026-11-17 |
Microsoft |
New in November 2026. |
| xAI | grok-4.6 |
1 |
xAI |
|
| xAI | grok-4-1-fast-reasoning |
1 |
xAI |
|
| xAI | grok-4 |
1 |
xAI |
|
| DeepSeek | DeepSeek-V4.1-Flash |
2026-09-10 |
DeepSeek |
New in November 2026. |
| DeepSeek | DeepSeek-V3.2 |
1 |
DeepSeek |
|
| Meta | Llama-4-Maverick-17B-128E-Instruct-FP8 |
1 |
Meta |
|
| Fireworks | FW-GLM-5.3 |
1 |
Fireworks |
Enable Fireworks first. See Fireworks models. |
| Fireworks | FW-GLM-5.3-Flash |
1 |
Fireworks |
Enable Fireworks first. See Fireworks models. |
| Fireworks | FW-Kimi-K3 |
1 |
Fireworks |
Enable Fireworks first. See Fireworks models. |
Claude models
A model router deployment routes to a Claude model only when both of these conditions are true:
- The Claude model is deployed in the same Foundry resource as your model router deployment, as a Global Standard deployment of the version in the preceding table. Model router doesn't use other versions, such as the Azure-hosted version
2that the Foundry portal selects by default. To deploy one, see Deploy and use Claude models. - If your model router deployment uses a model subset, the subset includes the Claude model.
These Claude versions are available only as Global Standard deployments, so a Data Zone Standard model router deployment doesn't route to Claude models.
Auto can route to a Claude model when your resource has a deployment of that model. Only Claude models answer Messages API requests.
In a request, set model to your model router deployment name or to auto. Setting model to a Claude model name doesn't select that model through model router. For the steps, see Add Claude models.
Fireworks models
Model router routes to the Fireworks models (FW-GLM-5.3, FW-GLM-5.3-Flash, and FW-Kimi-K3) only when Fireworks is enabled for your Azure subscription. To enable it, a subscription admin reviews the terms and registers the Fireworks.EnableDeploy preview feature. Without it, model router doesn't route to these models. For steps, see Enable Fireworks on Foundry.
How the model pool changes
New models are added to version 2025-11-18 without changing the version name. Retired models are removed. What's new in model router lists each change.
- If your deployment uses the full pool, new models are included and retired models drop out with no action from you.
- If your deployment uses a model subset, new models aren't added until you add them. If a model in your subset retires, replace it with a supported model. See Route to a subset of models.
List the current model pool programmatically
The Foundry models API returns the current pool for each model router version in a region. The following Azure CLI command returns the format, name, and version of each model in the version 2025-11-18 pool for East US 2:
SUBSCRIPTION_ID=$(az account show --query id --output tsv)
az rest --method get \
--url "https://management.azure.com/subscriptions/$SUBSCRIPTION_ID/providers/Microsoft.CognitiveServices/locations/eastus2/models?api-version=2025-06-01" \
--query "value[?model.name=='model-router' && model.version=='2025-11-18'] | [0].model.capabilities.models" \
--output tsv
The command returns a JSON array with one entry for each model. The output starts like this example:
[{"format":"OpenAI","name":"gpt-4o","version":"2024-11-20"},{"format":"OpenAI","name":"gpt-4o-mini","version":"2024-07-18"}, ...]
Supported regions
Model router version 2025-11-18 supports the following regions. A check mark (✅) means that the deployment type is available. A hyphen (-) means that it isn't. Auto (model router without a deployment) is available in all 32 regions and always processes data as Global Standard.
| Region | Global Standard | Data Zone Standard |
|---|---|---|
| Australia East | ✅ | ✅ |
| Brazil South | ✅ | - |
| Canada Central | ✅ | - |
| Canada East | ✅ | - |
| Central US | ✅ | ✅ |
| East US | ✅ | ✅ |
| East US 2 | ✅ | ✅ |
| France Central | ✅ | ✅ |
| Germany West Central | ✅ | ✅ |
| Italy North | ✅ | ✅ |
| Japan East | ✅ | ✅ |
| Japan West | ✅ | ✅ |
| Korea Central | ✅ | ✅ |
| North Central US | ✅ | ✅ |
| North Europe | ✅ | ✅ |
| Norway East | ✅ | ✅ |
| Poland Central | ✅ | ✅ |
| South Africa North | ✅ | - |
| South Central US | ✅ | ✅ |
| South India | ✅ | ✅ |
| Southeast Asia | ✅ | ✅ |
| Spain Central | ✅ | ✅ |
| Sweden Central | ✅ | ✅ |
| Switzerland North | ✅ | ✅ |
| Switzerland West | ✅ | - |
| UAE North | ✅ | - |
| UK South | ✅ | - |
| UK West | ✅ | - |
| West Central US | ✅ | - |
| West Europe | ✅ | ✅ |
| West US | ✅ | ✅ |
| West US 3 | ✅ | ✅ |
In each region, model router routes only across the supported models that are available in that region.
- Global Standard: prompts and responses can be processed in any geography where the selected model is deployed.
- Data Zone Standard: prompts and responses are processed within the data zone of your resource.
For details, see Data, privacy, and security for Foundry Models sold by Azure. Model router isn't available in Azure Government. For more information, see Foundry in Azure Government.
Versions and lifecycle
Model router uses dated versions. The latest version keeps receiving new models without a version change. Older versions keep a fixed set of models.
| Version | Status | Retirement date | Model pool |
|---|---|---|---|
2025-11-18 |
Generally available; default for new deployments | May 20, 2027 | Updated over time |
2025-08-07 |
Preview | January 27, 2027 | Fixed |
2025-05-19 |
Preview | January 27, 2027 | Fixed |
A retirement date applies to that model router version, not to model router. Move deployments on 2025-05-19 or 2025-08-07 to 2025-11-18 before January 27, 2027. For the steps, see Update the version.
To check a deployment's version update setting, see Model updates. For all Foundry model retirement dates, see Foundry Models lifecycle and support policy.