Skip to content
POST/api/v1/prompts/ab-experiments

Create a prompt A/B experiment

Creates an experiment + 2-8 variants in one call. The worker runs each variant against the dataset with the configured scorers and declares a winner via significance testing. requiredRole: editor.

Authentication

Send Authorization: Bearer YOUR_API_KEY on every request. Generate API keys at /dashboard/api-keys.

Request body required

Example

{
  "projectId": "00000000-0000-0000-0000-000000000000",
  "name": "string",
  "description": "string",
  "prompt_name": "string",
  "model": "gpt-4o-mini",
  "dataset_id": "00000000-0000-0000-0000-000000000000",
  "scorer_ids": [
    "string"
  ],
  "variants": [
    {
      "label": "string",
      "content": "string",
      "weight": 1
    }
  ],
  "winner_config": {
    "metric": "score",
    "direction": "higher_is_better",
    "confidence": 0.95,
    "min_samples": 30
  }
}
Schema
{
  "application/json": {
    "schema": {
      "type": "object",
      "properties": {
        "projectId": {
          "type": "string",
          "format": "uuid",
          "pattern": "^([0-9a-fA-F]{8}-[0-9a-fA-F]{4}-[1-8][0-9a-fA-F]{3}-[89abAB][0-9a-fA-F]{3}-[0-9a-fA-F]{12}|00000000-0000-0000-0000-000000000000|ffffffff-ffff-ffff-ffff-ffffffffffff)$"
        },
        "name": {
          "type": "string",
          "minLength": 1,
          "maxLength": 200
        },
        "description": {
          "type": "string",
          "maxLength": 2000
        },
        "prompt_name": {
          "type": "string",
          "minLength": 1,
          "maxLength": 200
        },
        "model": {
          "default": "gpt-4o-mini",
          "type": "string",
          "minLength": 1,
          "maxLength": 64
        },
        "dataset_id": {
          "type": "string",
          "format": "uuid",
          "pattern": "^([0-9a-fA-F]{8}-[0-9a-fA-F]{4}-[1-8][0-9a-fA-F]{3}-[89abAB][0-9a-fA-F]{3}-[0-9a-fA-F]{12}|00000000-0000-0000-0000-000000000000|ffffffff-ffff-ffff-ffff-ffffffffffff)$"
        },
        "scorer_ids": {
          "minItems": 1,
          "maxItems": 20,
          "type": "array",
          "items": {
            "type": "string"
          }
        },
        "variants": {
          "minItems": 2,
          "maxItems": 8,
          "type": "array",
          "items": {
            "type": "object",
            "properties": {
              "label": {
                "type": "string",
                "minLength": 1,
                "maxLength": 64
              },
              "content": {
                "type": "string",
                "minLength": 1,
                "maxLength": 50000
              },
              "weight": {
                "default": 1,
                "type": "integer",
                "minimum": 1,
                "maximum": 100
              }
            },
            "required": [
              "label",
              "content",
              "weight"
            ],
            "additionalProperties": false
          }
        },
        "winner_config": {
          "type": "object",
          "properties": {
            "metric": {
              "default": "score",
              "type": "string",
              "enum": [
                "score",
                "pass_rate",
                "latency",
                "cost"
              ]
            },
            "direction": {
              "default": "higher_is_better",
              "type": "string",
              "enum": [
                "higher_is_better",
                "lower_is_better"
              ]
            },
            "confidence": {
              "default": 0.95,
              "type": "number",
              "minimum": 0.5,
              "maximum": 0.999
            },
            "min_samples": {
              "default": 30,
              "type": "integer",
              "minimum": 10,
              "maximum": 10000
            }
          },
          "required": [
            "metric",
            "direction",
            "confidence",
            "min_samples"
          ],
          "additionalProperties": false
        }
      },
      "required": [
        "projectId",
        "name",
        "prompt_name",
        "model",
        "scorer_ids",
        "variants"
      ],
      "additionalProperties": false
    }
  }
}

Response

201 example

{
  "success": true
}

All status codes

201Experiment created ({ id, status:'draft', variants }).
400(no description)
401(no description)
403Forbidden — editor role required.
404Project not found.
429(no description)

Code samples

cURL

curl -X POST \
  https://evalguard.ai/api/v1/prompts/ab-experiments \
  -H "Authorization: Bearer $EVALGUARD_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{ "projectId": "00000000-0000-0000-0000-000000000000", "name": "string", "description": "string", "prompt_name": "string", "model": "gpt-4o-mini", "dataset_id": "00000000-0000-0000-0000-000000000000", "scorer_ids": [ "string" ], "variants": [ { "label": "string", "content": "string", "weight": 1 } ], "winner_config": { "metric": "score", "direction": "higher_is_better", "confidence": 0.95, "min_samples": 30 } }'

TypeScript

import { EvalGuard } from "@evalguard/sdk";

const client = new EvalGuard({ apiKey: process.env.EVALGUARD_API_KEY });

const response = await client.request({
  method: "POST",
  path: "/api/v1/prompts/ab-experiments",
  body: {
    "projectId": "00000000-0000-0000-0000-000000000000",
    "name": "string",
    "description": "string",
    "prompt_name": "string",
    "model": "gpt-4o-mini",
    "dataset_id": "00000000-0000-0000-0000-000000000000",
    "scorer_ids": [
      "string"
    ],
    "variants": [
      {
        "label": "string",
        "content": "string",
        "weight": 1
      }
    ],
    "winner_config": {
      "metric": "score",
      "direction": "higher_is_better",
      "confidence": 0.95,
      "min_samples": 30
    }
  },
});
console.log(response);

Python

from evalguard import EvalGuard
import os

client = EvalGuard(api_key=os.environ["EVALGUARD_API_KEY"])

response = client.request(
    method="POST",
    path="/api/v1/prompts/ab-experiments",
    body={
    "projectId": "00000000-0000-0000-0000-000000000000",
    "name": "string",
    "description": "string",
    "prompt_name": "string",
    "model": "gpt-4o-mini",
    "dataset_id": "00000000-0000-0000-0000-000000000000",
    "scorer_ids": [
        "string"
    ],
    "variants": [
        {
            "label": "string",
            "content": "string",
            "weight": 1
        }
    ],
    "winner_config": {
        "metric": "score",
        "direction": "higher_is_better",
        "confidence": 0.95,
        "min_samples": 30
    }
},
)
print(response)

Go

package main

import (
	"context"
	"fmt"
	"net/http"
	"os"
	"strings"
)

func main() {
	body := strings.NewReader(`{"projectId":"00000000-0000-0000-0000-000000000000","name":"string","description":"string","prompt_name":"string","model":"gpt-4o-mini","dataset_id":"00000000-0000-0000-0000-000000000000","scorer_ids":["string"],"variants":[{"label":"string","content":"string","weight":1}],"winner_config":{"metric":"score","direction":"higher_is_better","confidence":0.95,"min_samples":30}}`)
	req, _ := http.NewRequestWithContext(context.Background(), "POST", "https://evalguard.ai/api/v1/prompts/ab-experiments", body)
	req.Header.Set("Authorization", "Bearer "+os.Getenv("EVALGUARD_API_KEY"))
	req.Header.Set("Content-Type", "application/json")
	resp, err := http.DefaultClient.Do(req)
	if err != nil { panic(err) }
	defer resp.Body.Close()
	fmt.Println(resp.Status)
}

Errors

400401403404429

Other Prompts endpoints