POST/api/v1/generate-eval-suite

Generate eval suite from natural language

Accepts a plain-English description of what to test and generates a complete eval configuration with test cases, scorers, and thresholds.

Authentication

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

Request body required

Example

{
  "description": "Test a customer support chatbot that handles returns, refunds, and order tracking for an e-commerce platform",
  "output_format": "yaml",
  "test_density": "standard"
}
Schema
{
  "application/json": {
    "schema": {
      "type": "object",
      "required": [
        "description"
      ],
      "properties": {
        "description": {
          "type": "string",
          "minLength": 10,
          "maxLength": 10000,
          "description": "Natural language description of the eval suite to generate",
          "example": "Test a customer support chatbot that handles returns, refunds, and order tracking for an e-commerce platform"
        },
        "output_format": {
          "type": "string",
          "enum": [
            "yaml",
            "json"
          ],
          "default": "yaml",
          "description": "Output format for the generated config"
        },
        "test_density": {
          "type": "string",
          "enum": [
            "light",
            "standard",
            "comprehensive"
          ],
          "default": "standard",
          "description": "How many test cases to generate"
        }
      }
    }
  }
}

Response

200 example

{
  "success": false,
  "data": {
    "config": {},
    "metadata": {},
    "usage_logs": [
      {}
    ]
  }
}

All status codes

200Generated eval suite
400(no description)
502Pipeline execution failed

Code samples

cURL

curl -X POST \
  https://evalguard.ai/api/v1/generate-eval-suite \
  -H "Authorization: Bearer $EVALGUARD_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{ "description": "Test a customer support chatbot that handles returns, refunds, and order tracking for an e-commerce platform", "output_format": "yaml", "test_density": "standard" }'

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/generate-eval-suite",
  body: {
    "description": "Test a customer support chatbot that handles returns, refunds, and order tracking for an e-commerce platform",
    "output_format": "yaml",
    "test_density": "standard"
  },
});
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/generate-eval-suite",
    body={
    "description": "Test a customer support chatbot that handles returns, refunds, and order tracking for an e-commerce platform",
    "output_format": "yaml",
    "test_density": "standard"
},
)
print(response)

Go

package main

import (
	"context"
	"fmt"
	"os"

	"github.com/evalguard/evalguard-go"
)

func main() {
	client := evalguard.NewClient(os.Getenv("EVALGUARD_API_KEY"))
	resp, err := client.Request(context.Background(), "POST", "/api/v1/generate-eval-suite", map[string]any{"description": "Test a customer support chatbot that handles returns, refunds, and order tracking for an e-commerce platform", "output_format": "yaml", "test_density": "standard"})
	if err != nil { panic(err) }
	fmt.Println(resp)
}

Errors

400502

Other NL Pipeline endpoints