Cohere
by Cohere Inc.
Generate text with Command models, create embeddings for semantic search, rerank search results, classify text, and summarise documents using Cohere's NLP API.
How agents use Cohere
- ✓Agent embeds user queries and document chunks for RAG retrieval using Cohere's embed-english model
- ✓Agent reranks vector search results to improve relevance before passing context to an LLM
- ✓Agent classifies incoming support messages by category using few-shot examples
- ✓Agent summarises long documents or articles before forwarding them to downstream nodes
- ✓Agent uses Command-R+ to reason over retrieved context and generate grounded responses
Agent actions
Inputs: message, model, systemPrompt, chatHistory, temperature, maxTokens
Returns: text, finishReason, inputTokens, outputTokens
Inputs: texts, model, inputType
Returns: embeddings, embeddingCount, model
Inputs: query, documents, topN, model
Returns: results, topDocument, topScore
Inputs: inputs, examples
Returns: classifications, labels
Inputs: text, length, format, extractiveness
Returns: summary
Example workflows
RAG pipeline with reranking
Embed a query, retrieve candidates from Pinecone, rerank with Cohere, then generate a response with Command-R+.
About Cohere
- Vendor
- Cohere Inc.
- Pricing Always review details with the vendor
- Free / Paid — Free trial includes limited API calls. Production pricing is usage-based per token/request.
- Authentication
- API key
- Rate limit Always review details with the vendor
- 100 requests / minute
- Compatible nodes
- AgentResource
- Website
- https://cohere.com
Put Cohere behind your AI agent
Set up Cohere alongside your other tools in the planner — no code — so AI can act on it for your customers.
Open the planner