n8n-nodes-cognidata

n8n community node that integrates with Cognidata Suite — a unified AI gateway with token tracking and cost control.

What is Cognidata Suite?

Cognidata Suite acts as a centralized gateway for LLM providers (OpenAI, Anthropic, Google, Groq, Ollama, etc.). Instead of configuring each provider separately, you route all AI requests through Cognidata Suite, which handles:

• Unified access to multiple LLM providers with a single API key
• Token tracking and cost control per user
• Model permissions — administrators control which models each user can access
• Agent Flows — pre-built AI workflows that can be triggered via API

Installation

In n8n (Community Nodes)

1. Go to Settings > Community Nodes
2. Enter n8n-nodes-cognidata
3. Click Install

Via npm

npm install n8n-nodes-cognidata

Via Docker

A ready-to-use Docker setup is included in the docker/ directory:

cp docker/.env.example docker/.env
# Edit docker/.env with your Cognidata API key
cd docker && docker compose up --build

Setup

1. In n8n, go to Credentials > New Credential > Cognidata Suite API
2. Enter your API Key (format: sk-cog-...)
3. Enter the Base URL of your Cognidata Suite instance (e.g. http://localhost:8000/api/v1)
4. Click Save — the credential will validate by fetching your available models

Node: Cognidata AI

Chat Completion

Send messages to any LLM model available in your Cognidata Suite instance.

| Parameter | Description | |-----------|-------------| | Model | Dropdown that dynamically loads your available models | | System Prompt | Optional system instructions for the AI | | User Message | The message to send (supports n8n expressions) | | Temperature | Controls randomness (0.0 – 2.0, default: 0.7) | | Max Tokens | Maximum tokens in the response (default: 1024) |

Output:

| Field | Description | |-------|-------------| | content | The AI model's response text | | model_id | UUID of the model used | | model_name | Display name of the model | | usage | Token usage (input_tokens, output_tokens, total_tokens) | | finish_reason | Why the model stopped (e.g. stop) |

Trigger Flow

Execute a pre-configured Cognidata Suite Agent Flow.

| Parameter | Description | |-----------|-------------| | Flow ID | The UUID of the Agent Flow to trigger | | Input Message | Input text for the flow |

API Endpoints Used

| Operation | Method | Endpoint | |-----------|--------|----------| | Credential test | GET | /inference/models | | Load models | GET | /inference/models | | Chat completion | POST | /inference/chat | | Trigger flow | POST | /inference/flows/{flowId}/execute |

License

MIT