Chat completions
Create a completion for the provided prompt and parameters. Compatible with OpenAI’s chat completions API. Supports 40+ models including GPT-4, Claude, Gemini, and CaseMark legal AI models. Includes streaming support, token counting, and usage tracking.
Authorizations
API key starting with sk_case_
Headers
Stable internal user or customer identifier for AI Gateway cost reporting.
256Comma-separated AI Gateway reporting tags. At most 10 unique tags, each 1–64 characters.
Body
List of messages comprising the conversation
Model to use for completion. Defaults to casemark/core-large if not specified
"casemark/core-large"
Maximum number of tokens to generate
1000
Sampling temperature between 0 and 2
0.7
Whether to stream back partial progress
false
CaseMark-only: controls whether reasoning fields appear in responses. Defaults to false (suppressed) for most CaseMark models; defaults to true for casemark/core-potassium.
false
Nucleus sampling parameter
Frequency penalty parameter
Presence penalty parameter

