GitButler Logo
API Reference

Evaluate Prompt

Operations about evaluate_prompts

POST
/evaluate_prompt/predict

Post Evaluate Prompt Predict

Evaluate an AI Prompt

/evaluate_prompt/predict

Request Body

application/jsonRequired

messages[role]
Required
array<string>

messages[content]
Required
array<string>

max_tokensinteger

Maximum number of output tokens, maximum 400

Default: 300Format: "int32"

temperaturenumber

How creative the response should be. Between 0 and 2, the lower the less creative

Format: "float"

systemstring

For Anthropic, set the system prompt to use

model_kindstring

Which model provider should be used

Default: "openai"Value in: "openai" | "anthropic"
curl -X POST "https://app.gitbutler.com/api/evaluate_prompt/predict" \
  -H "Content-Type: application/json" \
  -d '{
    "messages[role]": [
      "string"
    ],
    "messages[content]": [
      "string"
    ],
    "max_tokens": 300,
    "temperature": 0.1,
    "system": "string",
    "model_kind": "openai"
  }'

Evaluate an AI Prompt