SentioGate FAQ
General
What is SentioGate?
SentioGate is an AI LLM router: one OpenAI-compatible API (https://api.sentiogate.dev/v1) in front of many model providers. Requests made with a SentioGate gateway key run on your own provider keys at 0% token markup, or on the free community pool until you add keys. With model "auto" it picks a capable model for each request and automatically falls back to another model if one fails.
How do I start using SentioGate?
Sign up with an email address, create a gateway key (it starts with sg_live_) in the Vault, then point any OpenAI-compatible client at https://api.sentiogate.dev/v1 with that key and send model "auto". No credit card is needed to start.
Is SentioGate hosted or self-hosted?
Hosted only. SentioGate runs on a global edge network, so there are no servers, containers or databases for you to run. If you must run the gateway inside your own infrastructure, a self-hosted gateway such as LiteLLM is the better fit.
How do I contact SentioGate?
Email support@sentiogate.app for help with your account or the API, or sales@sentiogate.app for enterprise and volume questions.
Why choose SentioGate as an OpenRouter alternative?
For 0% token markup on your own provider keys. You can still attach an OpenRouter key inside SentioGate for wider model coverage.
Why choose SentioGate as a LiteLLM alternative?
You get a hosted OpenAI-compatible endpoint with nothing to run. If you must self-host, LiteLLM is the better pick.
What is an LLM router?
A layer between your app and model providers: you call one API, and it picks the model for each request and tries another when one fails.
What is the difference between an LLM gateway and an LLM router?
A router chooses the model or provider per request; a gateway is the wider control layer that also covers keys, limits, logs and policy.
Free tier
Is there a free chatbot API that does not need a credit card?
Yes. Create a free account and a gateway key, then send model "auto" to the free community pool: 600 requests per 5 hours per account.
What are the limits of the free community pool?
600 requests per rolling 5-hour window per account. The pool is shared, so at busy times a request can get HTTP 429 with a Retry-After header.
Which model does the free chatbot API use?
Send model "auto". SentioGate chooses the model behind the free community pool, and it can change over time.
What happens when I use up the free community pool?
Requests return HTTP 429 with code community_pool_exhausted and a Retry-After header. Add your own provider key to keep going.
Which LLM APIs are free in 2026?
Several providers have free tiers or free model variants. The list checks each one on its official page.
Pricing & BYOK
Which providers can I connect to SentioGate?
The Vault has built-in support for OpenAI, Anthropic, Google Gemini, Groq, OpenRouter, Mistral, xAI and DeepSeek keys. You can also add your own OpenAI-compatible endpoint, such as a vLLM or Ollama server, as a custom provider.
Is SentioGate free?
Yes. The Free plan costs $0 during the 90-day public beta and needs no credit card.
Does SentioGate mark up token prices?
No. On your own keys each provider bills you directly, and SentioGate adds 0% token markup.
What will Pro and Team cost?
Pro is planned at $9/month and Team at $18/month. Both are coming soon and cannot be bought yet.
What does BYOK mean?
Bring your own key: your provider keys are stored in the Vault, encrypted at rest, and requests run on your own provider accounts.
What happens when I add my own API keys?
Your account switches to BYOK routing, the free community pool turns off for it, and requests run on your keys at 0% token markup.
How many keys can I add per provider?
Up to 5 keys per provider, in priority order.
Routing & failover
Does SentioGate support streaming responses?
Yes. Send stream: true and the provider’s server-sent events stream is passed through in the OpenAI format, so the streaming helpers in the OpenAI SDKs work unchanged. Router details for a streamed request arrive in the X-SentioGate-Router response header.
Can SentioGate stop an agent that is stuck in a loop?
Yes. When an agent looks stuck repeating the same steps, SentioGate flags it in the X-SentioGate-Loop response header and, by default, answers with a 400 loop_detected error instead of calling the model. Send X-SentioGate-Loop-Stop: off, or turn off "Stop runaway loops" in the dashboard, to only flag it.
How does SentioGate handle OpenAI 429 rate limits and provider outages?
With model "auto" it routes around models that are failing and automatically falls back to another model if one fails.
How do I avoid 429 rate limit errors when building chatbots on free AI APIs?
Add free-tier keys from several providers to the Vault and send model "auto". SentioGate honors Retry-After and moves on to your next key or model.
How does SentioGate reduce LLM token bills without changing application code?
Cost-first routing with model "auto", trimming long conversations, response caching and per-key spend caps.
Integrations
Can I send SentioGate request traces to Langfuse?
Yes, as an opt-in. Connect your own Langfuse project and SentioGate exports request traces to it, with metadata only by default. The Langfuse guide covers the setup.
Is the free AI API compatible with the OpenAI SDK?
Yes. Point the official openai Python or Node.js library at https://api.sentiogate.dev/v1 with your gateway key.
Can I use this free chatbot API for Discord, Slack, or conversational AI bots?
Yes. Any bot framework that accepts an OpenAI-compatible base URL works.
Can Claude Code use SentioGate?
Yes, through the Anthropic Messages endpoint: set ANTHROPIC_BASE_URL to the API host without /v1 and use your SentioGate key.
How do I point OpenCode at SentioGate?
Add a custom provider to opencode.json that uses @ai-sdk/openai-compatible with the SentioGate base URL.
Updated · By Hisham, founder of SentioGate