Overview
Credentials store sensitive API keys and authentication tokens for third-party services (OpenAI, Pinecone, etc.). All credential data is encrypted at rest.All credential endpoints require JWT authentication. See Authentication for details.
List All Credentials
Retrieve all credentials in your workspace.GET /api/v1/credentials
Query Parameters
Filter credentials by type (e.g.,
openAIApi, pineconeApi)Response
Unique identifier for the credential
Display name for the credential
Type of credential (e.g.,
openAIApi, pineconeApi, huggingFaceApi)Encrypted credential data (API keys, secrets, etc.)
ID of the workspace this credential belongs to
Credential values are encrypted and never returned in plain text through the API.
Get Credential by ID
Retrieve a specific credential by its ID.GET /api/v1/credentials/:id
Path Parameters
The unique identifier of the credential
Response
Returns a single credential object with the same structure as shown in the list response.Create Credential
Create a new credential for a third-party service.POST /api/v1/credentials
Request Body
Display name for the credential
Type of credential. Common values:
openAIApi- OpenAI APIpineconeApi- Pinecone vector databasehuggingFaceApi- HuggingFacecohereApi- CohereanthropicApi- Anthropic ClaudegoogleGenAIApi- Google Generative AI
Plain-text credential data (will be encrypted before storage)Structure varies by credential type:
Response
Update Credential
Update an existing credential.PUT /api/v1/credentials/:id
Path Parameters
The unique identifier of the credential to update
Request Body
Include only the fields you want to update:Updated display name
Updated credential data (will be re-encrypted)
Response
Returns the updated credential object.Delete Credential
Delete a credential permanently.DELETE /api/v1/credentials/:id
Path Parameters
The unique identifier of the credential to delete
Response
Credential Types
Flowise supports credentials for numerous integrations:LLM Providers
LLM Providers
- OpenAI (
openAIApi) - GPT models - Anthropic (
anthropicApi) - Claude models - Cohere (
cohereApi) - Cohere models - Google (
googleGenAIApi) - Gemini models - HuggingFace (
huggingFaceApi) - Open-source models
Vector Databases
Vector Databases
- Pinecone (
pineconeApi) - Qdrant (
qdrantApi) - Weaviate (
weaviateApi) - Milvus (
milvusApi) - Chroma (
chromaApi)
Document Stores
Document Stores
- Airtable (
airtableApi) - Notion (
notionApi) - Confluence (
confluenceApi)
Other Services
Other Services
- Serper (
serperApi) - Google Search - SerpAPI (
serpApi) - Search results - Replicate (
replicateApi) - Model hosting
Code Examples
Security Considerations
Encryption at Rest
All credentials are encrypted using AES-256-GCM before storage
Workspace Isolation
Credentials are scoped to workspaces and cannot be accessed cross-workspace
No Plain Text Retrieval
Encrypted data is never returned in plain text through the API
Audit Trail
Created and updated timestamps track credential lifecycle
Best Practices
- Use descriptive names - Make it easy to identify credentials at a glance
- One credential per environment - Separate dev, staging, and production credentials
- Regular rotation - Update credentials periodically for security
- Delete unused credentials - Remove credentials that are no longer needed
- Check dependencies - Before deleting, verify no chatflows depend on the credential