Find the model for your workload
Check current model availability and choose the route you will actually test.
Open the model catalogRun your first request
Choose a model, create an API key, and call CometAPI
Follow a short, testable path from model selection to a verified response, with focused guides for contract and error checks.
request-check.md
1
POST /v1/chat/completions2
verify auth header3
confirm model route4
record fallback behavior1Choose a current model
2Create a scoped API key
3Run and verify a request
Three practical entry points
Move from model choice to a verified CometAPI request
Start with the task blocking you now, then use the contract and error checks before sending production traffic.
Run a contract smoke test
Confirm the endpoint, response fields, finish reason, and usage shape.
Use the request checklistRead an error without guessing
Separate status, request shape, model selection, and retry decisions.
Open the error workflowLatest Guides
Recent source-checked guides
Connect Zapier to CometAPI for No-Code AI Responses
Use Zapier’s native CometAPI action to turn mapped trigger data into a tested AI response workflow.
Build an AI Workflow with CometAPI and Activepieces
Connect Activepieces to CometAPI with its native Ask CometAPI action, test a controlled prompt, and operate the flow with safe logging and clear recovery steps.
Build a Streaming Streamlit Chat App with CometAPI
Build a Python Streamlit chat interface that sends conversation history to CometAPI and renders incremental assistant responses.
Build a Haystack RAG Pipeline with CometAPI
Configure Haystack’s in-memory retrieval pipeline with local embeddings and CometAPI chat generation, then test and operate it safely.
Connect goose to CometAPI with a Custom Provider
Configure goose with CometAPI, choose a tool-capable model, and verify chat and file operations safely.