mintoken
Start free →
Getting started

Quickstart

From zero to compressed API response in about two minutes. We'll create an account, generate an API key, and make your first call from Python, TypeScript, or cURL.

1. Create a mintoken account

Head to mintoken.in/signup and sign up with an email and password. No credit card, no email verification — you land on the dashboard immediately with the free plan active (100,000 tokens per month).

2. Create an API key

From the dashboard, navigate to API Keys and click New key. Give it a name and pick a compression intensity:

  • lite — ~40% savings. Conservative; keeps articles and natural-sounding prose. Good default for customer-facing output.
  • full — ~65% savings. Drops filler words, fragment-style. Good for technical / internal workloads. This is the default.
  • ultra — ~75% savings. Maximally abbreviated, uses arrows and symbols for causality. Best for logs, pipelines, agent-to-agent messages.
Copy it immediately
The raw API key is shown exactly once, at creation time. Mintoken stores only a SHA-256 hash — we can't recover it later. If you lose it, rotate the key and grab the new one.

What a mintoken key looks like

All keys start with mt_live_ followed by 64 hexadecimal characters. Example: mt_live_9c59e9081b12db9c5c79…. Treat it like a password — do not commit it to source, do not paste it in Slack, do not put it in client-side code.

3. Make your first call

The mintoken API is a drop-in replacement for the OpenAI API — you use the OpenAI SDK exactly as you always would, but with two tweaks:

  • Change base_url to https://api.mintoken.in/v1
  • Set api_key to your mt_live_… key
  • Add a new X-Provider-Key header carrying your OpenAI key
from openai import OpenAI

client = OpenAI(
    base_url="https://api.mintoken.in/v1",
    api_key="mt_live_xxxxx",                      # your mintoken key
    default_headers={
        "X-Provider-Key": "sk-proj-...",          # your own OpenAI key
    },
)

response = client.chat.completions.create(
    model="gpt-4o-mini",
    messages=[{"role": "user", "content": "Explain connection pooling"}],
)

print(response.choices[0].message.content)
print(f"Tokens: {response.usage.completion_tokens}")

Run it. You'll see a compressed response — probably ~30-50 tokens for a three-sentence answer that would normally cost 80-150 tokens. In the response headers, look for X-Mintoken-Intensity: full to confirm mintoken ran on the request.

Why two keys?
The mt_live_ key identifies you on mintoken — we use it to track your usage, enforce quotas, and scope your analytics. The X-Provider-Key is your OpenAI / Anthropic / Google key — mintoken uses it to make the upstream call on your behalf. We never store it.

4. Store keys in environment variables

Don't hard-code either key into your source files. Put them in .env (and make sure .env is in your .gitignore):

MINTOKEN_API_KEY=mt_live_xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
OPENAI_API_KEY=sk-proj-...

Then load them in your app — with python-dotenv, dotenv, or your framework's built-in env handling.

5. Verify the savings

After a few calls, head to /dashboard. You'll see:

  • The total dollars saved this month (estimated against a no-mintoken baseline).
  • Total tokens compressed and request count.
  • A usage bar showing how much of your plan quota you've used.

If you prefer numbers in code, call the analytics endpoint — see the Analytics page.

What's next