OpenAI
Detailed information on the OpenAI conversation component
Component format
A Dapr conversation.yaml component file has the following structure:
apiVersion: dapr.io/v1alpha1
kind: Component
metadata:
name: openai
spec:
type: conversation.openai
metadata:
- name: key
value: mykey
- name: model
value: gpt-4-turbo
- name: endpoint
value: 'https://api.openai.com/v1'
- name: responseCacheTTL
value: 10m
# - name: apiType # Optional
# value: 'azure'
# - name: apiVersion # Optional
# value: '2025-01-01-preview'
Warning
The above example uses secrets as plain strings. It is recommended to use a secret store for the secrets, as described here.Spec metadata fields
| Field | Required | Details | Example |
|---|---|---|---|
key | Y | API key for OpenAI. | mykey |
model | N | The OpenAI LLM to use. Defaults to gpt-5-nano, or to the OPENAI_MODEL environment variable when it is set. | gpt-4-turbo |
maxTokens | N | Default maximum number of tokens the model may generate per request. A request-level maxTokens value overrides this default. When unset, no cap is sent and the provider’s own default applies. | 2048 |
useLegacyMaxTokensField | N | Send the token cap as max_tokens instead of max_completion_tokens. Set this when endpoint points at an OpenAI-compatible service that doesn’t recognize max_completion_tokens, otherwise the cap is ignored. Defaults to false. | true |
endpoint | N | Custom API endpoint URL for OpenAI API-compatible services. If not specified, the default OpenAI API endpoint is used. Required when apiType is set to azure. | https://api.openai.com/v1, https://example.openai.azure.com/ |
responseCacheTTL | N | Time-to-live for the in-memory response cache. When set, identical requests are served from cache until they expire. | 10m |
apiType | N | Specifies the API provider type. Required when using a provider that does not follow the default OpenAI API endpoint conventions. | azure |
apiVersion | N | The API version to use. Required when the apiType is set to azure. | 2025-04-01-preview |
useResponsesAPI | N | Send requests to the OpenAI Responses API (/v1/responses) instead of the Chat Completions API. Set this to use function tools with reasoning models that reject tools on Chat Completions. Requests are sent with store: false, so OpenAI keeps no conversation state. Not supported when apiType is set to azure: the component fails to initialize. Defaults to false. | true |
Azure OpenAI Configuration
To configure the OpenAI component to connect to Azure OpenAI, you need to set the following metadata fields which are required for Azure’s API format.
Required fields for Azure OpenAI
When connecting to Azure OpenAI, the following fields are required:
apiType: Must be set toazureto enable Azure OpenAI compatibilityendpoint: Your Azure OpenAI resource endpoint URL (e.g.,https://your-resource.openai.azure.com/)apiVersion: The API version for your Azure OpenAI deployment (e.g.,2025-01-01-preview)key: Your Azure OpenAI API key
Get your configuration values from: https://ai.azure.com/
Azure OpenAI component example
apiVersion: dapr.io/v1alpha1
kind: Component
metadata:
name: azure-openai
spec:
type: conversation.openai
metadata:
- name: key
value: "your-azure-openai-api-key"
- name: model
value: "gpt-4.1-nano" # Default: gpt-4.1-nano
- name: endpoint
value: "https://your-resource.openai.azure.com/"
- name: apiType
value: "azure"
- name: apiVersion
value: "2025-01-01-preview"
Note
When using Azure OpenAI, bothendpoint and apiVersion are mandatory fields. The component returns an error if either field is missing when apiType is set to azure.