What to Ask Leina to Do with Cohere
Copy a prompt, add your details, and send it to Leina in your connected team chat.
Use source documents as context for a focused question.
Use Cohere model [model ID] to answer [question] with documents [documents]. Ask for citations where supported and explain unanswered points. Keep source-based statements distinguishable from general explanation.
Find the most relevant candidates within a supplied document set.
Use Cohere rerank model [model ID] for query [query] and candidate documents [documents], returning the top [N]. Map result indices back to the original document titles and show relevance scores as relative ranking signals, not truth or probability measures.
Get Started with Cohere
-
Connect Your Account
Connect with a Cohere API key.
-
Try a Task
Tell Leina what you need in your team chat. Include the source, destination, and expected result.
Set Up Leina in Your Chat → -
Check the Result
Review the result and any changes in the connected app.
Supported Cohere Capabilities
Generate chat responses with supplied documents, rerank document candidates for a query, and create text embeddings through Cohere. These operations work with the material you provide; they do not create a searchable knowledge base on their own.
- Answers and relevance
-
Generate document-aware answers and rerank supplied candidates.
- Chat
- Rerank Documents
- Text embeddings
-
Create vectors for supplied text with the appropriate input type.
- Embed Texts
View All 3 Actions
- Chat
- Embed Texts
- Rerank Documents
Frequently Asked Questions
What documents should I provide for an answer?
Provide relevant text and identifying titles or IDs. A focused document set makes it easier to connect statements to their sources.
How should I use reranking results?
Use the indices and relevance scores to prioritize review of the supplied candidates. Reranking does not verify factual accuracy or search outside that set.
Can I create embeddings for search?
Yes, supply a compatible model, texts and input type appropriate to documents or queries. Store and search those vectors through a separate retrieval system.
Your AI employee for Cohere
Tell Leina what you need in your team chat. Include the source, destination, and expected result.