One key.
Pay for what you use.
Reach Claude, Gemini and DeepSeek models through one URL. Claude Code, Cursor and your own code connect to the same URL. Each request is deducted from your balance by token count, and you can see every cent.
POST /v1/messages # model "wiozen-anthropic"
What does a request cost?
| Example request | Input tokens | From cache | Output tokens | Cost |
|---|---|---|---|---|
| Coding agent: file edit | 14,820 | 83,150 | 1,940 | $0.001545 |
| Coding agent: short step | 2,310 | 96,480 | 620 | $0.000722 |
| One-off chat question | 1,870 | 0 | 1,120 | $0.000318 |
| Coding agent: long reply | 21,400 | 71,900 | 3,480 | $0.002126 |
Cost = token count × price per 1M tokens. Token counts are examples. Your own request breakdown appears in this format in your dashboard.
Your first request in three steps.
Create an account
Verify your email and $0.15 in trial credit is added right away. No credit card required.
Create a key
Create a separate key for each tool in the dashboard, and optionally set a spend limit per key.
Set the URL in your tool
In Claude Code, Cursor or your own code, change the URL and key. Keep using the tool the way you already do.
No surprises.
The gateway's job is simple: get your request to a working provider and show you exactly what you spent.
One key, two protocols
The same sk-gw- key works with both the Anthropic and OpenAI formats. Streaming, tool calls and token counting are supported.
Pay as you go
No subscriptions or bundles. Your USD balance drops by the actual token cost of each request, and topped-up balance never expires.
Falls back when a provider fails
When an upstream provider returns an error or a rate limit, the request moves to the next provider. Failed requests are not charged.
A record of every request
Input, cache and output tokens, along with the USD cost, appear request by request in your dashboard.
No daily limit
As long as your balance lasts, you can keep sending requests. When it runs out, the API returns 429. Top up and you pick up where you left off.
Invite and earn 10%
10% of your invitee's first top-up goes to both of you as bonus credit (your invite link). If you get stuck, open a support ticket from the dashboard.
Top up, spend per token.
No bundles, no multipliers. Add any amount and it becomes USD balance. Each request is deducted at the prices below. Cached input is priced at one tenth.
wiozen-anthropicDeal- Input / 1M
- $0.05
- Cache / 1M
- $0.005
- Output / 1M
- $0.20
claude-opus-5-5−%30- Input / 1M
$4.00$2.80- Cache / 1M
- $0.14
- Output / 1M
$20.00$14.00
claude-sonnet-5-5−%30- Input / 1M
$2.00$1.40- Cache / 1M
- $0.14
- Output / 1M
$10.00$7.00
gemini-3.8-flash-high−%60- Input / 1M
$0.75$0.30- Cache / 1M
- $0.03
- Output / 1M
$3.75$1.50
deepseek-v4.1-flash−%50- Input / 1M
$0.30$0.15- Cache / 1M
- $0.003
- Output / 1M
$1.20$0.60
Create your account and verify your email. Trial credit is added right away. No credit card is required.
Create account- agent requests
- –
- chat messages
- –
- validity
- no expiry
Minimum $5. Payment is by bank transfer and is added to your balance once confirmed.
Connect in two lines.
Set the URL and your key, and keep using the tool you already know. The Anthropic format runs through /v1/messages, and the OpenAI format through /v1/chat/completions.
export ANTHROPIC_BASE_URL="https://api.wiozen.com" export ANTHROPIC_AUTH_TOKEN="sk-gw-..." export ANTHROPIC_MODEL="wiozen-anthropic" export ANTHROPIC_DEFAULT_HAIKU_MODEL="wiozen-anthropic" claude
curl https://api.wiozen.com/v1/chat/completions \ -H "Authorization: Bearer sk-gw-..." \ -H "Content-Type: application/json" \ -d '{"model": "wiozen-anthropic", "messages": [{"role": "user", "content": "Hello"}]}'
from openai import OpenAI client = OpenAI(base_url="https://api.wiozen.com/v1", api_key="sk-gw-...") reply = client.chat.completions.create( model="wiozen-anthropic", messages=[{"role": "user", "content": "Hello"}], )
Short answers.
What is Wiozen API?
Wiozen is an OpenAI- and Anthropic-compatible API gateway that gives you access to AI models through one URL and one key. You use it by changing only the URL and key in your existing code.
Which models are available?
Currently live: Wiozen Anthropic, Claude Opus 5.5, Claude Sonnet 5.5, Claude Haiku 5.5, Gemini 3.8 Flash, DeepSeek V4.1 Flash. The current list and prices per 1M tokens are on the Models page.
How does pricing work?
There is no subscription. You top up your balance and pay only for the tokens you use. The minimum top-up is $5. Topped-up balance never expires, and the cost of each request is shown separately in your dashboard.
Can I try it for free?
Yes. New accounts get $0.15 in trial credit. No credit card is required. Trial credit is given once per account.
Which tools does it work with?
It works with Cline, Cursor, Roo Code, Kilo Code, Continue, Claude Code, Hermes Agent, OpenCode, Aider, Qwen Code, Open WebUI and LibreChat, as well as the OpenAI and Anthropic SDKs. Each tool has a step-by-step setup guide.
Is uptime guaranteed?
No. Some models are served through third-party providers, so outages, slowdowns or model changes can happen. When a provider returns an error, the request is routed to a backup provider, and failed requests are not charged.
How does the referral reward work?
Share your invite link from your dashboard. Once the first top-up made by someone who signed up with it is confirmed, 10% of that amount is added to both of your accounts as bonus credit.
How do I get support?
Open a ticket from the Support section of your dashboard. We will let you know by email when we reply.