One Gateway
150+ models across 22 providers
OpenAI-compatible AI Gateway
One gateway for every model
Route every AI request through one Base URL and one API key. Switch providers, control cost, and keep logs without rewriting your app.
Smart Routing
Route by availability, latency, and policy
Cost Visibility
Track requests, tokens, failures, and spend
Key Control
Centralize API keys and provider access
- OpenAI-compatible API
- Policy-based routing
- Logs, tokens, and cost
Change one endpoint. Keep your stack.
Use the OpenAI-compatible format you already know. Replace the provider base URL, keep your SDK, and let route the request.
- One Base URL
- OpenAI-compatible
- Logs and cost tracking
Create a gateway key
Open Console and create an API key for your app, IDE, or agent runtime.
Replace the Base URL
Point your existing SDK or tool to the Neotoken endpoint below.
env
base_url = "https://api.neotoken.io/v1"
api_key = "mr_your_key_here"Send and observe
Send your first request. Routing, fallback, usage, and logs stay in one place.
bash
curl https://api.neotoken.io/v1/chat/completions \
-H "Authorization: Bearer mr_your_key_here" \
-H "Content-Type: application/json" \
-d '{"model":"gpt-5.5","messages":[{"role":"user","content":"Hello"}]}'Connect your tools
Cursor, WebStorm / JetBrains, OpenAI SDK, and curl — OpenAI-compatible, minimal migration cost.
Cursor
Configure the BaseURL and API key in Cursor AI settings to access all models managed by .
View Cursor guideCursor Settings
// Cursor → Settings → AI → OpenAI API Key
{
"openai.apiKey": "mr_your_key_here",
"openai.baseUrl": "https://api.neotoken.io/v1"
}Control every route from Console
Manage gateway keys, model access, usage, and request logs from one self-serve workspace.
- Gateway KeysCreate scoped keys for apps, IDEs, and agents. Plaintext is shown once and safely copyable.
- Model AccessBrowse available models, provider coverage, and recommended access paths.
- Cost and UsageTrack tokens, requests, spend, and model share trends with flexible time filters.
- Routing LogsInspect timestamp, model, tokens, status, latency, and failure reason for every request.
Observe cost, latency, and reliability
Every request becomes measurable: tokens, spend, failures, model distribution, and routing health.
Requests1.2M+
Token Usage18B+
Success Rate99.4%
Cost¥2,840
Model Share
gpt-5.542%
claude-sonnet-4-631%
gpt-5.4-mini18%
gemini-3.5-flash9%
Get started today
For individual developers, teams, and enterprises.