All systems operational

Trox,
on your terms.

One key. Both protocols. Pay only for what you use. No subscriptions, no rate-limit gymnastics, no surprise bills. Live in 30 seconds.

Start free → Read the docs
1M context window · OpenAI compatible · Anthropic compatible · Streaming
# Replace base_url and api_key. Same SDK, different provider.
from openai import OpenAI

client = OpenAI(
    base_url="https://ai.qubith.in/v1",
    api_key="qbk_your_key_here",
)

reply = client.chat.completions.create(
    model="trox-prime",
    messages=[{"role": "user", "content": "Say hi in 5 words."}],
    max_tokens=64,
)

print(reply.choices[0].message.content)

Three reasons people switch.

Why build a backend for one model when the gateway already exists?

Faster than free

Sign up, paste your card, ship in 30 seconds. No infra to stand up, no SDK to write, no model to evaluate.

One key, both protocols

Use the same key with OpenAI's SDK or Anthropic's. Pick whichever your codebase already uses — no rewrites.

Pay only for what you use

Tokens never expire. No monthly minimum. Spin up a side project without a $20/mo subscription holding you hostage.

Simple, transparent pricing.

Three tiers. No tiers hidden behind a sales call.

Starter
₹29 / once
29,000 tokens · ≈25 chats
  • All models, both APIs
  • OpenAI + Anthropic compatible
  • Streaming
  • Tokens never expire
Get started
Power
₹299 / once
2,000,000 tokens · ≈1,800 chats
  • Everything in Pro
  • Production-grade volume
  • ~25× the tokens of Pro
  • Per-call output cap: 8k tokens
Get started

Need more? Contact us for volume pricing.

Ready to ship?

Free to sign up. Buy a package when you're ready. No subscription, no email gate, no demo call.

Create your account →