Route to your favorite providers
Why Napsix AI Gateway?
Built for developers who need transparent billing, support in their own language, and peace of mind knowing exactly where every token went.
One API, hundreds of models
OpenAI-compatible endpoint. Swap the baseURL and you're done — your existing code with the OpenAI SDK works without any changes.
Smart routing
Automatic multi-key load balancing. We drain credits nearing expiration and fail over in milliseconds if a provider goes down.
BYOK free forever
Bring your own Anthropic / OpenAI / AWS keys. You only pay the subscription for features — no markup on your usage.
Per-request observability
Every call is audited with request_id, latency, time-to-first-token, cost in microdollars, and which real key ran it.
Native SSE streaming
Full chat.completion.chunk support with time-to-first-token tracking. Tools, function calling and vision included.
SDK + documentation
Compatible with the OpenAI SDK, Vercel AI SDK, LangChain, LlamaIndex and any OpenAI-compatible client.
Transparent pricing
BYOK always free. You pay for features and included tokens.
Free
$0 /mo
- 100,000 tokens a month
- 5 API keys
- Basic models (Llama 3.3, DeepSeek, GPT-OSS)
- No credit card
Starter
$99 /mo
- 10 million tokens a month
- 25 API keys
- Basic and premium models (Sonnet, GPT-5-mini)
- Bring your own Anthropic or OpenAI keys
- 14-day trial
Pro
$499 /mo
- 100 million tokens a month
- Unlimited API keys
- Basic, premium and frontier (Opus, GPT-5, Gemini 2.5 Pro)
- Your own keys + priority routing
- 14-day trial
Enterprise
Custom
- Unlimited tokens and keys
- SLA defined in your contract, plus dedicated support
- Your own keys, priority routing and white-label
- Endpoint on your own domain
