Sonar Deep Research — exhaustive research & comprehensive reports
sonar-deep-research
Note: Deep research requests may take significantly longer to complete (up to several minutes) due to the exhaustive multi-step search process.
citations and search_results fields. Deep research responses are typically much longer and more detailed than other models.Bearer authentication header of the form Bearer <token>, where <token> is your auth token.
The Sonar model to use.
sonar, sonar-pro, sonar-reasoning-pro, sonar-deep-research A list of messages comprising the conversation so far.
The maximum number of tokens to generate in the response.
Sampling temperature between 0 and 2. Lower values make output more focused and deterministic.
0 <= x <= 2Nucleus sampling parameter. The model considers tokens with top_p probability mass.
0 <= x <= 1The number of tokens to keep for top-k filtering.
0 <= x <= 2048Whether to stream the response using server-sent events.
Controls how much search context to use. Affects per-request cost.
low, medium, high Penalizes new tokens based on their existing frequency in the text so far. Positive values decrease the likelihood of repeating the same line verbatim.
0 <= x <= 2Penalizes new tokens based on whether they appear in the text so far. Positive values increase the likelihood of talking about new topics.
-2 <= x <= 2Whether to return citations and search results in the response.
Filter search results by recency.
month, week, day, hour Limit search to specific domains.
Successful response with AI answer and citations
Unique identifier for the completion.
The model used for the completion.
"chat.completion"
Unix timestamp of when the completion was created.
List of source URLs referenced in the answer.
Detailed search results with titles, snippets, and URLs.