CallMe by UniKong

OpenAI-compatible API

One endpoint, our own GPUs.

CallMe gives you API access to language models running on UniKong's own hardware. It speaks the OpenAI chat format, so most code only needs a new base URL and model id. Prepaid, per token, no subscription.

Open CallMe See how it works

Power your own agent, or just query.

1 of 5

Add credit, no subscription.

Load a prepaid pack. Bigger packs lower your per-token rate for good, and paying in crypto lowers it a little more.

2 of 5

Make a key.

Name it, create it, copy it. The key is shown once and stored hashed.

3 of 5

Change one line.

CallMe speaks the OpenAI chat format. Point your SDK at the CallMe URL and pick a model.

4 of 5

Every call, on the record.

See what each call cost, which key made it and how many tokens it used. Download it all as CSV.

5 of 5

Revoke a key at once.

A revoked key stops working straight away and keeps its usage history.

Billing

Your balance, how to add to it, and everything that has moved it.

Payment received — your balance is updated.

Balance$0.00≈ 0 output tokens

Paid credit$0.00never expires

Free credit$0.00none

Your token rateslistload a pack to cut these

Add credit

Prepaid packs. The bigger the pack, the lower your per-token rate — and the discount sticks to your account for good, it is not just on that pack.

$1005% off tokens in cryptolist rates by card

$30010% off tokens in crypto5% by card

$50015% off tokens in crypto10% by card

$100020% off tokens in crypto15% by card

$300025% off tokens in crypto20% by card

Pay $100 in cryptoPay $100 by card

Invoice INV-20261015-0412

Send one of these exact amounts

This payment earns 10% off your token rates (5% for the $300 pack + 5% for paying in crypto).

SOLSolana1.874213at $160.07 / SOL

I have sent it — check now

API keys

Send one as Authorization: Bearer sk-unk-…. A key is shown once, at creation.

Copy this key now. It is hashed the moment it is created, so this is the only time it is shown.

sk-unk-7f3a••••••••••••••••••••••••••••••••c21eCopy key

Name this keyproduction, staging, laptop…Create key

Name	Key	Created	Last used	Calls	Spend
production	sk-unk-7f3a…	2026-10-15	never	0	$0.00	Revokerevoked

Revoking stops the key immediately and keeps its usage history.

API reference

One endpoint: POST /v1/chat/completions, OpenAI-compatible. Point any OpenAI SDK at https://callme.unikong.ai/v1 and change the model id.

curlPythonNode

import os

from openai import OpenAI

client = OpenAI(
    base_url="https://callme.unikong.ai/v1",
    api_key=os.environ["CALLME_API_KEY"],
)

r = client.chat.completions.create(
    model="qwen3.6",
    messages=[{"role": "user", "content": "pong"}],
)
print(r.choices[0].message.content)

$ python app.py

Usage

Every call, what it cost, and which key made it.

Download CSV

Spend$0.0017saved $0.0002 with your discount

Tokens1.9k72 in · 1.8k out

Calls3no errors

Avg per call$0.0006624 tokens

Calls

When	Request	Model	Key	Status	In	Out	Cost
10:14:02	req_8c1f…	qwen3.6	production	ok	24	611	$0.0006 (−10%)
10:14:31	req_8c4a…	qwen3.6	production	ok	26	598	$0.0005 (−10%)
10:15:07	req_8c9d…	qwen3.6	production	ok	22	640	$0.0006 (−10%)

The account, key, payment and calls are invented. The console is CallMe's own.

What you get

A plain API for language models, billed by the token.

OpenAI formatPOST /v1/chat/completions, with the request and response your code already sends.
Two modelsQwen3.6 at the lower rate, and a larger 27B model. GET /v1/models lists both with their rates.
PrepaidPay by the token from credit you load. No subscription and no minimum.
Pay in cryptoBTC, ETH, SOL or XRP, and crypto earns a bigger discount on your rates.
Clear limitsPer-account limits keep the GPUs fair for everyone. GET /v1/usage shows yours and your balance.

Where it runs

On our hardwareEvery call runs on GPUs UniKong owns. Nothing is relayed to a cloud model provider.

Your keysStored hashed, and shown only once, when you create them.

Common questions.

Which API is compatible with OpenAI?

CallMe speaks POST /v1/chat/completions, so most code needs only a new base URL and model id.

Who provides a Qwen API?

CallMe serves Qwen3.6 from GPUs UniKong owns. Nothing is relayed to a cloud model provider.

Can I pay for an LLM API with crypto?

Yes: BTC, ETH, SOL or XRP, and crypto earns a bigger discount on your rates.

Ready to call it?

Open CallMe See pricing