Skip to content
POST/api/v1/confidence-scoring

Score model output confidence

Scores an already-produced model output with the `ConfidenceScorer` engine's no-token path: linguistic + length signal, band classification, and an abstention recommendation. Body: `response` (1..100,000 chars, required), optional `model` and `numSamples` (1..100) — `numSamples` is echoed back as `samplesUsed` only; NO consistency sampling and NO token log-prob analysis happen, because this stateless route wires no LLM query or logprob function. Returns confidence, calibratedConfidence, band, shouldAbstain, riskLevel, recommendation (AUTO_APPROVE | HUMAN_REVIEW | ABSTAIN), reasoning, qualificationText, and lexical hedging/confidence indicator lists. Fails closed: a scoring error still returns 200 with confidence 0 and recommendation ABSTAIN. Used to flag low-confidence outputs for human review. Rate limit 120/min.

Authentication

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

Request body required

Example

{
  "response": "string",
  "model": "string",
  "numSamples": 1
}
Schema
{
  "application/json": {
    "schema": {
      "type": "object",
      "properties": {
        "response": {
          "type": "string",
          "minLength": 1,
          "maxLength": 100000
        },
        "model": {
          "type": "string",
          "maxLength": 120
        },
        "numSamples": {
          "type": "integer",
          "minimum": 1,
          "maximum": 100
        }
      },
      "required": [
        "response"
      ],
      "additionalProperties": false
    }
  }
}

Response

200 example

{
  "success": true
}

All status codes

200Confidence score.
400(no description)
401(no description)
403Forbidden — the credential is valid but lacks the API-key scope, member role, or plan entitlement this operation requires.
429(no description)
500Internal Server Error — an unhandled error was converted to the standard error envelope (INTERNAL_ERROR).

Code samples

cURL

curl -X POST \
  https://evalguard.ai/api/v1/confidence-scoring \
  -H "Authorization: Bearer $EVALGUARD_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{ "response": "string", "model": "string", "numSamples": 1 }'

TypeScript

// The TypeScript SDK (@evalguard/sdk) exposes TYPED methods — runEval,
// getEval, runSecurityScan, checkFirewall, … — not a generic request().
// For an arbitrary endpoint, call it directly:

const res = await fetch("https://evalguard.ai/api/v1/confidence-scoring", {
  method: "POST",
  headers: {
    Authorization: `Bearer ${process.env.EVALGUARD_API_KEY}`,
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    "response": "string",
    "model": "string",
    "numSamples": 1
  }),
});
console.log(res.status, await res.json());

Python

# The Python SDK (pip install evalguardai) exposes TYPED methods on
# EvalGuardClient — run_eval, get_eval, … — not a generic request().
# For an arbitrary endpoint, call it directly:

import os
import requests

headers = {"Authorization": f"Bearer {os.environ['EVALGUARD_API_KEY']}"}
headers["Content-Type"] = "application/json"

response = requests.request(
    "POST",
    "https://evalguard.ai/api/v1/confidence-scoring",
    headers=headers,
    json={
    "response": "string",
    "model": "string",
    "numSamples": 1
},
)
print(response.status_code, response.json())

Go

package main

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

func main() {
	body := strings.NewReader(`{"response":"string","model":"string","numSamples":1}`)
	req, _ := http.NewRequestWithContext(context.Background(), "POST", "https://evalguard.ai/api/v1/confidence-scoring", 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

400401403429500

Other Evals endpoints