AI APIs & SDKs

Labels

Which three fields are required in every Anthropic Messages API request body?

2 months ago

In the Anthropic Messages API, what are the two valid roles for message objects in the messages array?

2 months ago

Where should you place a system prompt when calling the Anthropic Messages API?

2 months ago

What does the max_tokens parameter control in the Anthropic Messages API?

2 months ago

What does the stop_sequences parameter do in the Messages API?

2 months ago

How are Claude model tiers broadly characterised?

2 months ago

How do you specify which Claude model to use in a Messages API request?

2 months ago

Why is it recommended to pin a specific model version (e.g. "claude-sonnet-4-5") in production rather than using an alias like "claude-sonnet-latest"?

2 months ago

You need to classify thousands of short customer support emails in real time. Which Claude model tier is most appropriate?

2 months ago

What is streaming in the context of the Anthropic API?

2 months ago

Why does streaming improve perceived latency even though total generation time is unchanged?

2 months ago

How do you enable streaming when using the official Anthropic Python SDK?

2 months ago

In which scenario is streaming NOT recommended?

2 months ago

In Anthropic's pricing model, which tokens are more expensive per unit?

2 months ago

What cache_control value do you attach to a content block to enable prompt caching?

2 months ago

When does prompt caching deliver the greatest cost savings?

2 months ago

What is the purpose of the Anthropic token counting endpoint?

2 months ago

What is the primary advantage of the Message Batches API over individual Messages API calls?

2 months ago

How do you provide tool definitions to the Anthropic Messages API?

2 months ago

How does the API signal that the model wants to call a tool?

2 months ago