Docs/Resources

Plan Limits

Per-plan request and token limits, response headers, and how to handle 429s.

Limits keep the platform fair and stable for everyone. Every plan has a monthly allowance. Your client should expect 429 responses and retry sensibly.

Plans

Plan	Price	Requests / month	Tokens / month	Seats
Free	$0	100	200K	1
Starter	$15/mo	1,000	1.5M	1
Pro	$49/mo	8,000	12M	1
Team	$149/mo	40,000	60M	5
Enterprise	Custom	Custom	Custom	Unlimited

Free plan: no credit card

Start right away with 100 requests and 200K tokens a month. All 60+ models are available on every paid plan. Free, Starter and Pro are individual plans; Team and Enterprise are separate plans for team workspaces, not an upgrade from Pro.

Two kinds of limit

Limit	Applies to	What you get
Monthly quota (table above)	You, or the team owner's plan	`429` with `code: rate_limit_exceeded` and no `Retry-After`. It resets at the start of next month.
Short bursts	Your client IP (not your plan)	`POST /v1/ai/`: 20 requests/minute. Everything else: 100/minute. `429` with* a `Retry-After` header.

Don't retry forever on a quota 429

If the 429 has code: rate_limit_exceeded and no Retry-After, you're out of monthly quota. Retrying won't help: back off for a long time or upgrade. A loop that only waits for Retry-After will spin.

Headers you can read

Burst-limited responses carry the standard headers:

HTTP

HTTP/1.1 200 OK
RateLimit-Limit: 100
RateLimit-Remaining: 95
RateLimit-Reset: 42          # seconds until the window resets
Retry-After: 42              # only on a 429

Successful chat requests also tell you where you stand on your monthly quota:

HTTP

X-Monthly-Requests-Used: 1240
X-Requests-Limit: 8000
X-Requests-Remaining: 6760
X-Monthly-Tokens-Used: 402113
X-Tokens-Limit: 12000000

Retry the right way

This honours Retry-After for bursts and uses capped exponential backoff with jitter otherwise:

JavaScript

async function makeRequestWithRetry(url, options, maxRetries = 4) {
  for (let attempt = 0; attempt <= maxRetries; attempt++) {
    const response = await fetch(url, options);
    if (response.status !== 429 && response.status < 500) return response;
    if (attempt === maxRetries) return response; // give up — let the caller see it

    const retryAfter = Number(response.headers.get('Retry-After'));
    const waitMs = Number.isFinite(retryAfter) && retryAfter > 0
      ? retryAfter * 1000                                              // burst: server told us
      : Math.min(2 ** attempt * 1000, 30000) + Math.random() * 1000;   // quota / 5xx: backoff + jitter

    await new Promise(r => setTimeout(r, waitMs));
  }
}

const response = await makeRequestWithRetry('https://api.tarqaai.com/api/v1/chat/completions', {
  method: 'POST',
  headers: { 'Authorization': 'Bearer YOUR_API_KEY', 'Content-Type': 'application/json' },
  body: JSON.stringify({
    model: 'gemini-2.5-flash',
    messages: [{ role: 'user', content: 'Hello!' }],
  }),
});

Tips

Watch the X-Requests-Remaining header and slow down before you hit zero.
Cache responses you'll ask for again.
Batch work where you can.
Need more? Contact support about higher limits.