One API for every frontier model
Use Claude, GPT, Gemini and more with one API key. Works with the OpenAI and Anthropic SDKs you already have, with usage analytics and simple token billing. Change the base URL, keep your code.
- Sign in with Google
- No credit card
- Free models included
fig. 1 — request
curl https://api.hexos.cloud/v1/chat/completions \ -H "Authorization: Bearer $HEXOS_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model": "claude-sonnet-4.6", "messages": [{"role": "user", "content": "Hello"}]}'terminal
fig. 2 — Response
Hello! How can I help you today?
- model
- claude-sonnet-4.6
- first token
- 380 ms
- tokens in → out
- 12 → 48
- status
- 200 OK
Send a request, get the answer streamed back. Every reply shows its model and tokens, so you always know what you were billed for.
Works with the model families you already use
- Claude
- GPT
- Gemini
- GLM
- Kimi
- DeepSeek
01 — Models
The models you already use, with one key
Call each model by its familiar name, like claude-sonnet-4.6 or gpt-6-astra. One account, one bill.
Claude
1.1GPT
1.2Gemini
1.3GLM
1.4Kimi
1.5DeepSeek
1.6
02 — Features
Built for production traffic
Everything you need to build on many models from one place.
2.1
Usage analytics
Per-key and per-model requests, tokens, latency and cache hit rate, with a full request log.
2.2
One key, every model
A single hxu- API key reaches Claude, GPT, Gemini, GLM, Kimi, DeepSeek and more. One account, one balance.
2.3
Reliable by default
Hexos keeps your requests flowing, so a busy model does not become a broken app.
2.4
Combos
List up to five models in order. Call combo/<name> and you get an answer from the first one that responds; you pay only for that one.
2.5
Drop-in compatible
OpenAI Chat Completions and Anthropic Messages. Point your SDK, Cursor, Cline or Claude Code at one base URL.
2.6
Playground
Try any model in the console before you write a line of code.
03 — How it works
From sign-in to first token in minutes
Sign in with Google
Open the console and continue with Google. No password to remember.
Create an API key
Generate an hxu- key on the API keys page. It is shown once, so store it safely.
Point your client at Hexos
Use any OpenAI or Anthropic client with the Hexos base URL. That's the whole migration.
https://api.hexos.cloud/v1
04 — Pricing
Simple token billing
Failed requests are free. Cached input costs a tenth of fresh input.
Free tier
4.0
Rp 0
A set of fast models is free for every account, with a daily request allowance.
See the free modelsStarter
4.1
Rp 10,000
For side projects and trying things out.
- tokens
- 20M
- validity
- 30 days
Pro
✓ Most popular
Rp 50,000
For daily coding agents and small teams.
- tokens
- 120M
- validity
- 30 days
Max
4.3
Rp 150,000
For heavy agent workloads and production apps.
- tokens
- 400M
- validity
- 30 days
Plans and redeem codes add tokens to your balance. Wallet top-ups are available in the console. Prices shown are current plans; see the console for the latest.
05 — FAQ
Questions, answered
5.1Is it compatible with my SDK?
Yes. Hexos speaks OpenAI Chat Completions and Anthropic Messages. Set the base URL and your hxu- key in the official SDKs, Cursor, Cline, Continue or Claude Code; nothing else changes.
5.2Which models are free?
A set of fast models is free for every account, with a daily request allowance. The current list is on the Free models page in the console.
5.3What if a model is busy or unavailable?
Hexos handles it for you and keeps serving the same model whenever it can. For extra safety, set up a combo so your request moves on to your next choice of model.
5.4Do failed requests cost tokens?
No. Only successful requests are billed. Cached input is billed at a tenth of fresh input.
5.5How do I pay?
Prices are in Indonesian rupiah (IDR). Buy a plan, top up your wallet in the console, or redeem a code; each adds tokens to your balance.
06 — /portal
Ship on every model today
Sign in with Google, create a key and send your first request in a couple of minutes.
Start building