API Documentation | MyAISpeed

Base URL

All API endpoints use the following base URL:

Base URL https://myaispeed-worker.vankristoff.workers.dev

All responses are JSON. No authentication is required for read endpoints. HTTPS is required for all requests.

List All Providers

GET /api/providers

Returns the full list of AI providers available for testing, including their display names, brand colors, supported models, and default model.

Auth: None required
Parameters: None

Response

{
  "providers": [
    {
      "slug": "openai",
      "name": "OpenAI",
      "color": "#10a37f",
      "models": ["gpt-4o", "gpt-4o-mini"],
      "defaultModel": "gpt-4o-mini"
    }
  ]
}

curl

curl https://myaispeed-worker.vankristoff.workers.dev/api/providers

JavaScript

const res = await fetch(
  "https://myaispeed-worker.vankristoff.workers.dev/api/providers"
);
const data = await res.json();
console.log(data.providers);

Python

import requests

res = requests.get(
    "https://myaispeed-worker.vankristoff.workers.dev/api/providers"
)
data = res.json()
print(data["providers"])

AI Speed Index

GET /api/index

Returns aggregated speed rankings for all AI providers based on real user measurements over a rolling 30-day window.

Auth: None required
Query params: country (optional) - ISO 3166-1 alpha-2 country code, e.g. US

Response

{
  "rankings": [
    {
      "provider": "openai",
      "avg_ttfb": 142.5,
      "avg_ttft": 310.2,
      "avg_tps": 48.7,
      "success_rate": 99.1,
      "test_count": 12450
    }
  ],
  "updated": "2026-03-08T12:00:00Z"
}

curl

# Global rankings
curl https://myaispeed-worker.vankristoff.workers.dev/api/index

# Filter by country
curl "https://myaispeed-worker.vankristoff.workers.dev/api/index?country=US"

JavaScript

const res = await fetch(
  "https://myaispeed-worker.vankristoff.workers.dev/api/index?country=US"
);
const data = await res.json();
console.log(data.rankings);

Python

import requests

res = requests.get(
    "https://myaispeed-worker.vankristoff.workers.dev/api/index",
    params={"country": "US"}
)
data = res.json()
print(data["rankings"])

City-Level Rankings

GET /api/index/cities

Returns speed rankings broken down by city. Useful for understanding regional performance differences.

Auth: None required
Query params: country (optional) - ISO 3166-1 alpha-2 country code

Response

{
  "cities": [
    {
      "city": "San Francisco",
      "country": "US",
      "provider": "openai",
      "avg_ttfb": 98.3,
      "test_count": 3200
    }
  ],
  "updated": "2026-03-08T12:00:00Z"
}

curl

# All cities
curl https://myaispeed-worker.vankristoff.workers.dev/api/index/cities

# Filter by country
curl "https://myaispeed-worker.vankristoff.workers.dev/api/index/cities?country=US"

JavaScript

const res = await fetch(
  "https://myaispeed-worker.vankristoff.workers.dev/api/index/cities?country=US"
);
const data = await res.json();
console.log(data.cities);

Python

import requests

res = requests.get(
    "https://myaispeed-worker.vankristoff.workers.dev/api/index/cities",
    params={"country": "US"}
)
data = res.json()
print(data["cities"])

Countries with Data

GET /api/index/countries

Returns a list of all countries that have test data, along with the number of tests from each country.

Auth: None required
Parameters: None

Response

{
  "countries": [
    { "country_code": "US", "cnt": 45200 },
    { "country_code": "GB", "cnt": 12800 },
    { "country_code": "DE", "cnt": 9400 }
  ]
}

curl

curl https://myaispeed-worker.vankristoff.workers.dev/api/index/countries

JavaScript

const res = await fetch(
  "https://myaispeed-worker.vankristoff.workers.dev/api/index/countries"
);
const data = await res.json();
console.log(data.countries);

Python

import requests

res = requests.get(
    "https://myaispeed-worker.vankristoff.workers.dev/api/index/countries"
)
data = res.json()
print(data["countries"])

Run a Connection Test

POST /api/test/connection

Runs a connection latency test against a specific AI provider. Measures DNS, TLS, and total round-trip time through the Cloudflare edge proxy.

Auth: None required
Rate limit: 5 requests per minute, 15 per hour
Content-Type: application/json

Request body

{
  "provider": "openai",
  "test_id": "550e8400-e29b-41d4-a716-446655440000"
}

Response

{
  "success": true,
  "provider": "openai",
  "ttfb_ms": 142,
  "dns_ms": 12,
  "tls_ms": 28,
  "total_ms": 156,
  "test_id": "550e8400-e29b-41d4-a716-446655440000"
}

curl

curl -X POST \
  https://myaispeed-worker.vankristoff.workers.dev/api/test/connection \
  -H "Content-Type: application/json" \
  -d '{"provider":"openai","test_id":"550e8400-e29b-41d4-a716-446655440000"}'

JavaScript

const res = await fetch(
  "https://myaispeed-worker.vankristoff.workers.dev/api/test/connection",
  {
    method: "POST",
    headers: { "Content-Type": "application/json" },
    body: JSON.stringify({
      provider: "openai",
      test_id: crypto.randomUUID()
    })
  }
);
const data = await res.json();
console.log(data.ttfb_ms, data.total_ms);

Python

import requests
import uuid

res = requests.post(
    "https://myaispeed-worker.vankristoff.workers.dev/api/test/connection",
    json={
        "provider": "openai",
        "test_id": str(uuid.uuid4())
    }
)
data = res.json()
print(f"TTFB: {data['ttfb_ms']}ms, Total: {data['total_ms']}ms")

CLI Tool

Run a speed test directly from your terminal with a single command. No installation required.

Usage

npx myaispeed

The CLI runs connection tests against all supported AI providers from your machine and prints a ranked table of results. It uses the same infrastructure as the web-based speed test.

Example output

  MyAISpeed - AI Provider Speed Test

  #  Provider       TTFB     Total
  1  Groq            48ms     62ms
  2  Cerebras        55ms     71ms
  3  OpenAI          98ms    142ms
  4  Anthropic      112ms    156ms
  5  Google         118ms    164ms
  ...

  Results submitted to AI Speed Index.

Test results are automatically submitted to the AI Speed Index to help build the global dataset. The same rate limits apply.

Rate Limits

To maintain data quality and prevent abuse, the following rate limits apply to the connection test endpoint:

Per-minute limit: 5 test requests per minute per IP address. Exceeding this returns a 429 Too Many Requests response.
Per-hour limit: 15 test requests per hour per IP address. This limit resets on a rolling window.
Read endpoints: The GET endpoints (/api/providers, /api/index, /api/index/cities, /api/index/countries) are not rate-limited but may be cached at the edge.

IP addresses are hashed with a salt before rate limit checks. Plaintext IPs are never stored.

Data Collection

Each test automatically collects anonymized metadata to power the AI Speed Index:

Geographic location (country, region, city) derived from Cloudflare's edge network
Network provider (ISP/ASN) from Cloudflare routing data
Device type (mobile, desktop, tablet) from User-Agent
Connection type (WiFi, 4G, 5G) from the Network Information API when available
Browser and OS from User-Agent

We do not collect personally identifiable information. IP addresses are never stored. All geographic and network data is derived from Cloudflare's edge infrastructure and stored in aggregate form only.

For full details, see the Methodology page.