n8n-nodes-caedral

Official n8n community node for Caedral (v1.0.0) — chat, vision, embeddings, voice, rerank, AI agents, and account usage through one prepaid API.

Installation

Self-hosted n8n

1. Open Settings → Community Nodes
2. Enter n8n-nodes-caedral
3. Click Install

Or via CLI inside your n8n data directory:

cd ~/.n8n
npm install n8n-nodes-caedral

Restart n8n after installation.

n8n Cloud

Community node verification for n8n Cloud is pending. Self-hosted instances can install from npm today.

Credentials

Create a Caedral API credential:

Setup walkthrough:

1. Sign up at caedral.com/signup
2. Open Dashboard → API Keys and create a key
3. In n8n, go to Credentials → Add credential → Caedral API
4. Paste your API key and save — n8n validates it with GET /v1/usage

API usage (except gated free Base / promo specialized) bills from prepaid balance. Top up at caedral.com/dashboard/billing.

Nodes

Caedral (main node)

Eight operations covering the full Caedral API:

Chat supports Simple mode (single message + optional system prompt) or JSON mode (full messages array).

Caedral Chat Model (AI sub-node)

Language model sub-node for n8n AI Agent and Chain nodes. Connect its Model output to the agent's language model input.

• Select chat tier (Base / Titan / Olympus / Primordial)
• Configure temperature and max tokens
• Works with all standard n8n AI Agent tool patterns

Caedral Embeddings (Vector Store sub-node)

Embedding model sub-node for n8n Vector Store nodes. Provides embedDocuments and embedQuery via POST /v1/embeddings.

• Default model: caedral-embed-e1-small-v1 (Caedral E1 Small, 384 dimensions)
• Legacy prepaid alias: caedral-embed (resolves to E1 Small on Caedral infrastructure)
• input_type: embedQuery sends query, embedDocuments sends document (E5-style prefixing)
• encoding_format: float (default) or base64 (decoded to float vectors for Vector Store nodes)
• Configurable batch size (default 512)

Caedral Reranker (Vector Store sub-node)

Reranker sub-node for Vector Store retrieval pipelines. Implements LangChain's compressDocuments via POST /v1/rerank.

• Default model: caedral-rerank
• Top N and minimum relevance score filters

Caedral Trigger (polling)

Polling trigger for account conditions (configure interval in n8n trigger settings):

Models & pricing

Authoritative pricing: caedral.com/pricing and caedral.com/models. All API usage bills from prepaid balance only.

Chat tiers

Specialized products

Example workflows

1. AI Agent with Caedral Olympus

1. Add an AI Agent node
2. Connect Caedral Chat Model as the Language Model input
3. Select Olympus for balanced agentic workloads
4. Attach tools (HTTP Request, Code, etc.) to the agent

2. RAG pipeline with embeddings and rerank

1. Trigger — new document arrives (webhook, schedule, etc.)
2. Caedral → Create Embedding — embed document chunks
3. Store vectors in your Vector Store node using Caedral Embeddings as the embedding model
4. On query: retrieve candidates, then pass through Caedral Reranker for relevance ordering
5. Feed top results to Caedral Chat Model or Chat Completion for the final answer

3. Low balance alert

1. Add Caedral Trigger → Balance Below Threshold
2. Set threshold to 1000 (= $10.00)
3. Connect to Slack, Email, or Discord notification node

4. Image generation webhook

1. Webhook trigger receives { "prompt": "..." }
2. Caedral → Generate Image with the prompt
3. Return the image URL or binary in the webhook response

Development

git clone https://github.com/trycaedral/n8n-nodes-caedral.git
cd n8n-nodes-caedral
npm install
npm run build
npm test

Run the official community scan (requires Node 22+ for the scanner's dependencies):

npx @n8n/scan-community-package n8n-nodes-caedral

Project structure

├── credentials/           # Caedral API credential type
├── nodes/
│   ├── Caedral/           # Main multi-operation node
│   ├── CaedralChatModel/  # AI Agent / Chain sub-node
│   ├── CaedralEmbeddings/ # Vector Store embeddings sub-node
│   ├── CaedralReranker/   # Vector Store reranker sub-node
│   └── CaedralTrigger/    # Polling trigger node
├── shared/                # Constants and pricing metadata
├── tests/                 # Integration tests
└── icons/                 # Node icons (light + dark)

Links

• Caedral Documentation
• API Reference
• n8n Integration Guide
• Pricing
• Report Issues

License

MIT