Max Tokens and Stop Sequences Quiz

Answers and explanations appear as you take the quiz.

1. What does max tokens limit?
   • The total tokens of input plus output in the context window
   • The number of tokens the model may generate in a single response
   • The number of requests per minute
   • The size of the model's vocabulary
2. Your call returns a finish reason of 'length' in OpenAI's API. What happened?
   • The model finished its answer naturally
   • A stop sequence was matched
   • The output hit the maximum token limit and was cut off
   • The model requested a tool call
3. Which Anthropic stop_reason indicates the model decided its answer was complete?
   • end_turn
   • max_tokens
   • stop_sequence
   • tool_use
4. Why does setting a low max tokens value not make answers concise?
   • The API ignores low values
   • The model writes as usual and is simply cut off when it reaches the limit
   • It makes the model skip the first half of its answer
   • It only affects input tokens
5. Is the matched stop sequence normally included in the returned text?
   • Yes, always at the end
   • Yes, but only when streaming
   • No, it is normally left out of the output
   • Only if it is a single character
6. A reasoning model returns an empty visible answer. What is a likely cause described in the guide?
   • The stop sequence was too long
   • Hidden reasoning tokens used up the output limit
   • The context window was too large
   • The prompt was written in lowercase
7. Which stop sequence is the best choice for ending after a structured answer block?
   • the
   • a single space
   • </answer>
   • .
8. When streaming a response, when should your code decide whether the output was truncated?
   • After the first chunk arrives
   • When the finish reason arrives at the end of the stream
   • Before sending the request
   • Streaming responses cannot be truncated