Simple, Transparent BYOK Pricing
Zero per-token markups. You pay model providers directly at raw API rates. During our 90-day beta, enjoy full adaptive model routing, context compression, and per-request cost tracking with measured cache savings at $0 platform fee. Paid plans are coming soon: tap Notify me and we'll email you when they launch.
Free Community Pool
5h Rolling WindowFree (90-Day Beta)
- Free during the 90-day public beta, no credit card
- 0% token markup: providers bill your own keys (BYOK) directly
- A free community pool (model "auto") while you have no keys
- Cost-first auto-routing across the providers you add
- Automatically falls back to another model if one fails
- Automatic context compaction for long sessions
- Runaway agent-loop stop, on by default
- Per-request cost tracking and measured cache savings
- Optional trace export to your own Langfuse project
ProComing soon
- Everything in Free
- No monthly request cap on your own keys
- Larger context limits before compaction
- Higher per-session token budgets
TeamComing soon
- Everything in Pro
- Higher per-session token budgets than Pro
- Planned: team seats with shared provider keys and settings
- Planned: team usage analytics, roles and an audit log
Why zero token markups?
Per-token markups misalign incentives with developers. With SentioGate, you connect your own provider API keys directly (BYOK) with 0% markup, paying raw list prices to providers. When paid plans launch, they will be a flat monthly fee for the routing layer, never a cut of your tokens.
Free community pool vs your own keys
| Free community pool | Your own keys (BYOK) | |
|---|---|---|
| Who pays for tokens | Nobody: free during the beta | Your providers bill you directly, 0% token markup |
| Limits | 600 requests per 5 hours per user | Your own provider rate limits |
| Models | model: "auto" only | auto across the models you enable (it automatically falls back to another model if one fails), or a pinned provider/model |
| Best for | Testing your integration | Production traffic |
How this compares with other ways to pay for LLM routing
- Credit-based routers such as OpenRouter bill tokens at the provider's list price and charge a fee when you buy credits (5.5% on its Standard plan, per its pricing page in October 2026).
- Self-hosted proxies such as LiteLLM have no gateway fee, but you run and secure the servers, the database and the upgrades yourself.
- SentioGate is hosted and charges no token markup on your own keys. It is free during the beta; paid plans are planned as a flat monthly fee.
For the details, see the OpenRouter vs direct API guide and the feature comparison.
Pricing FAQ
Is SentioGate free?
Yes. The Free plan costs $0 during the 90-day public beta and needs no credit card. Until you add provider keys you can use the free community pool with model "auto"; after that, requests run on your own keys (BYOK).
Does SentioGate mark up token prices?
No. With your own provider keys, each provider bills your account directly at its own prices, and SentioGate adds 0% token markup. When paid plans launch they will be a flat monthly fee for the routing layer, never a cut of your tokens.
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, so nothing is billed today. Use Notify me on a plan card to get one email when it launches.
What happens when I use up the free community pool?
Each account gets 600 requests per rolling 5-hour window. After that, requests return HTTP 429 with code community_pool_exhausted and a Retry-After header showing when your window frees up. To keep going, add your own provider key in the Vault.
What does BYOK mean?
Bring your own key. You save your own provider API keys (for example OpenAI, Gemini, Groq or OpenRouter) in the SentioGate Vault, encrypted at rest. Requests run on your provider accounts, so the provider bills you directly and your own rate limits apply. Adding keys turns the community pool off for your account.