Microsoft Foundry

Route FireConnect harnesses through Fireworks models deployed in your Azure subscription

FireConnect can route supported harnesses through a Fireworks model deployed in your Azure subscription. Azure bills this usage, and it counts toward your Microsoft Azure Consumption Commitment where applicable.

Before configuring FireConnect, open Microsoft Foundry to enable Fireworks and create a deployment.

CLI terminology: The Foundry provider is --provider azure, or --azure when you connect one harness. Harness configs display the label Fireworks on Microsoft Foundry. With Foundry, --model is your Azure deployment name, such as FW-GLM-5.2, not a Fireworks serverless ID such as glm-fast-latest.

Requirements

• A Microsoft Foundry resource with a Fireworks model deployment
• The resource endpoint and Azure API key
• FireConnect installed

See Coding Harnesses for current compatibility.

Use an Azure API key from Foundry, not a Fireworks key (fw_...).

Model routers are not available on the Foundry path. To use --model firerouter, first switch to the direct Fireworks gateway with fireconnect configure --provider fireworks.

Configure Foundry as the default

Set the Foundry endpoint once. New connections for supported harnesses use it until you change the default provider.

export AZURE_API_KEY="YOUR_AZURE_API_KEY"

fireconnect configure \
  --provider azure \
  --base-url "https://YOUR_RESOURCE.services.ai.azure.com"

If no Azure key is configured, FireConnect stores an {env:AZURE_API_KEY} reference. To store the current value literally, add:

fireconnect configure \
  --provider azure \
  --base-url "https://YOUR_RESOURCE.services.ai.azure.com" \
  --api-key "$AZURE_API_KEY"

Connect a harness

Use the deployment name from Foundry:

fireconnect <harness> --model <deployment-name>
fireconnect opencode --model FW-GLM-5.2

If you omit --model, FireConnect defaults to FW-GLM-5.2.

Accepted endpoint formats

Pass any of these Foundry URLs to --base-url. FireConnect converts it to https://<resource>.services.ai.azure.com/openai/v1:

• Bare resource root (https://<resource>.services.ai.azure.com)
• Portal project endpoint (.../api/projects/<name>)
• Foundry Models route (.../models)
• A complete OpenAI-compatible base URL (.../openai/v1)

Find the endpoint in the Microsoft Foundry portal under Project settings.

Route one command through Foundry

Route a single harness through Foundry without changing global config:

fireconnect opencode \
  --azure \
  --base-url "https://YOUR_RESOURCE.services.ai.azure.com" \
  --api-key $AZURE_API_KEY \
  --model FW-MiniMax-M2.5

If global config already has a Foundry endpoint, --azure alone reuses it:

fireconnect cursor --azure --model FW-GLM-5.2

FireConnect keeps Foundry configuration separate from direct Fireworks configuration. fireconnect <harness> status reports the provider, endpoint, and model. See Coding Harnesses for files changed and restore behavior.

fireconnect model list shows the Fireworks serverless catalog, not Foundry deployments. Enter a Foundry deployment name with --model.

Switch or disconnect

Goal	Commands
Return a supported harness to the Fireworks gateway	fireconnect configure --provider fireworks, then fireconnect <harness>
Remove FireConnect from one harness	fireconnect <harness> off
Restore all harnesses and remove FireConnect	fireconnect uninstall

You do not need to run off before reconnecting through Fireworks. The saved Azure endpoint and key remain available if you switch back later.

off does not change the global provider. If the provider remains azure, the next connection for a supported harness will use Foundry again.

Verify routing

fireconnect <harness> status

Confirm that the provider is azure and that the endpoint and deployment name are correct.

Related documentation

• Coding Harnesses
• Microsoft Foundry
• CLI Reference