Cloud provider configuration

Each cloud provider can add specified fields to providers.targets[]. This page shows the fields used for provider configuration and authentication.

Replace each your-...-model and deployment placeholder with an ID that is active for the provider account and region.

Azure OpenAI

providers:
  targets:
    - id: azure-primary
      provider: azure
      provider_type: azure-openai
      format: openai
      model: your-azure-openai-model
      base_url: https://replace-with-resource-name.openai.azure.com
      secret_key_ref:
        env: VERDICTAN_AZURE_OPENAI_API_KEY
      azure_api_version: your-supported-api-version
      azure_deployment: your-azure-deployment

Field	Type	Default	Notes
`base_url`	string	—	This field is necessary unless a different provider helper calculates it.
`azure_api_version`	string	legacy runtime fallback: `2024-02-01`	The Azure authentication builder sends this value. Set a version that your Azure deployment supports at this time. Do not use the fallback as your selected version.
`azure_deployment`	string	`model`	When this field is missing, the runtime uses the target `model`.
`path_template`	string	runtime/profile default	This optional field contains a specified request path template.

AWS Bedrock

providers:
  targets:
    - id: bedrock-claude
      provider: aws-bedrock
      provider_type: aws-bedrock
      format: anthropic
      model: anthropic.replace-with-bedrock-model-id
      base_url: https://bedrock-runtime.us-east-1.amazonaws.com
      aws_region: us-east-1
      bedrock_model_family: anthropic_messages

Field	Type	Default	Notes
`aws_region`	string	—	This necessary field gives the AWS region for Bedrock authentication signing. Set it in the provider target.
`bedrock_model_family`	string	—	Use `anthropic_messages` for the documented Anthropic-on-Bedrock request family.

aws_profile selects an optional profile in the default AWS credential chain. Bedrock authentication can use standard AWS profiles, workload credentials, and environment session credentials.

Keep secrets out of YAML. Use the standard AWS path that resolves credentials for the gateway process.

Bedrock targets use the AWS credential chain. A provider API key in secret_key_ref is not necessary.

When the signer runs, the specified AWS environment variables must be available.

Before rollout, verify the selected model family and request type in the configured region.

Google Vertex AI

providers:
  targets:
    - id: vertex-gemini
      provider: vertex-ai
      provider_type: google-vertex
      format: google-gemini
      model: your-gemini-model
      gcp_project: my-gcp-project
      gcp_region: us-central1

Field	Type	Default	Notes
`gcp_project`	string	—	This field is necessary for operation. It occurs in the Vertex endpoint, and lint gives a warning when it is missing.
`gcp_region`	string	`us-central1`	Vertex authentication uses this GCP region. Set it explicitly when your model runs in a different region.

Vertex targets use the configured Google authentication path. A provider API key in secret_key_ref is not necessary for the documented Vertex path.

Google AI Studio

providers:
  targets:
    - id: gemini-studio
      provider: google-ai-studio
      provider_type: google-ai-studio
      format: google-gemini
      model: your-gemini-model
      secret_key_ref:
        env: VERDICTAN_GEMINI_API_KEY

Google AI Studio uses the provider profile defaults. These defaults are x-goog-api-key and the google-gemini wire format.

No more cloud fields are necessary.

Anthropic

providers:
  targets:
    - id: anthropic-prod
      provider: anthropic
      provider_type: anthropic
      format: anthropic
      model: your-anthropic-model
      secret_key_ref:
        env: VERDICTAN_ANTHROPIC_API_KEY
      anthropic_version: 2023-06-01

Field	Type	Default	Notes
`anthropic_version`	string	`2023-06-01`	When the request does not contain this value, the gateway adds it as the `anthropic-version` header.
`api_key_header`	string	provider-profile default	Anthropic defaults to `x-api-key`.
`api_key_prefix`	string	provider-profile default	Anthropic defaults to an empty prefix.

Cloudflare AI

providers:
  targets:
    - id: cloudflare-llm
      provider: cloudflare-ai
      provider_type: cloudflare-ai
      format: openai
      model: your-cloudflare-model
      cloudflare_account_id: your-cloudflare-account-id
      secret_key_ref:
        env: VERDICTAN_CF_API_TOKEN

Field	Type	Notes
`cloudflare_account_id`	string	When `base_url` is missing, this field calculates `https://api.cloudflare.com/client/v4/accounts/{account_id}/ai/v1`.
`cloudflare_account_id_env`	string	This is the environment variable name that contains the account ID.

Snowflake Cortex

providers:
  targets:
    - id: snowflake-llm
      provider: snowflake-cortex
      model: your-snowflake-model
      snowflake_account_identifier: myorg-myaccount
      secret_key_ref:
        env: VERDICTAN_SNOWFLAKE_API_KEY

Field	Type	Notes
`snowflake_account_identifier`	string	When `base_url` is missing, this field calculates `https://{account_identifier}.snowflakecomputing.com`.
`snowflake_account_identifier_env`	string	This is the environment variable name that contains the account identifier.

Specified provider type

Use provider_type only when you must override the type from the provider profile.

providers:
  targets:
    - id: custom-openai-compatible
      provider: my-custom-provider
      provider_type: openai
      base_url: https://llm.example.com
      model: your-provider-model
      secret_key_ref:
        env: VERDICTAN_CUSTOM_PROVIDER_KEY

Supported specified values are:

openai
anthropic
cohere
huggingface
replicate
databricks
watsonx
aws-bedrock
google-ai-studio
google-vertex
sagemaker
azure-openai
cloudflare-ai
snowflake-cortex
generic

Wire format translation

providers:
  targets:
    - id: anthropic-explicit-format
      provider: anthropic
      provider_type: anthropic
      format: anthropic
      model: your-anthropic-model
      secret_key_ref:
        env: VERDICTAN_ANTHROPIC_API_KEY

Use format: anthropic when the target must use the native Anthropic wire contract. It does not replace provider fields or the tested workflow.

Supported format values are:

openai
anthropic
cohere
huggingface
replicate
watsonx
google-gemini

Other provider-level fields

Field	Type	Default	Notes
`allow_insecure_tls`	boolean	`false`	This field skips TLS verification for the target.
`region`	string	—	This field contains route data.
`weight`	number	—	Weighted routes use this value.
`timeout_seconds`	integer	`30`	This value sets the target timeout.
`stream_timeout_seconds`	integer	—	This field is an optional streaming override.
`quantizations`	`string[]`	—	Target values are `fp32`, `fp16`, `bf16`, `int8`, `int4`, `awq`, `gptq`, and `gguf`. Route filters accept a smaller set.

Complete example

pack:
  name: multi-cloud
  version: 1.0.0
  enabled: true
providers:
  targets:
  - id: azure-prod
    provider: azure
    provider_type: azure-openai
    format: openai
    model: your-azure-openai-model
    base_url: https://replace-with-resource-name.openai.azure.com
    secret_key_ref:
      env: VERDICTAN_AZURE_OPENAI_API_KEY
    azure_deployment: your-azure-deployment
    azure_api_version: your-supported-api-version
  - id: vertex-eu
    provider: vertex-ai
    provider_type: google-vertex
    format: google-gemini
    model: your-gemini-model
    gcp_project: your-gcp-project
    gcp_region: europe-west4
  - id: cloudflare-edge
    provider: cloudflare-ai
    provider_type: cloudflare-ai
    format: openai
    model: your-cloudflare-model
    cloudflare_account_id: your-cloudflare-account-id
    secret_key_ref:
      env: VERDICTAN_CF_API_TOKEN
  routing:
    strategy: ordered
policies:
  chain:
  - audit-logger
policy:
  audit-logger: {}

With ordered routing, the gateway uses its retry and provider-attempt rules to examine subsequent eligible targets.

If a deployment must stop on the first eligible target, use providers.routing.allow_fallbacks: false.

See provider attempts and fallback. Before you run the example, replace each placeholder with an account and region value.

Recommended use

Use the canonical snake_case provider fields.
Set provider_type only when you must override the provider profile type.
Make sure that format agrees with the wire protocol that the gateway must translate.
Use only the quantization values that are accepted at this time.
Do not use allow_insecure_tls for production traffic.

Azure OpenAI​

AWS Bedrock​

Google Vertex AI​

Google AI Studio​

Anthropic​

Cloudflare AI​

Snowflake Cortex​

Specified provider type​

Wire format translation​

Other provider-level fields​

Complete example​

Recommended use​

Next steps​