Pricing

Pay for passage, not for lock-in.

Metered by sealed request, with provider token costs passed through at cost. Start free, then turn on reputation routing, anchoring, and enclave verification as you scale.

Developer

$0/ forever

For local prototyping and your first integration tests.

Start free

10,000 inference requests / mo
Cryptographically signed passes
OpenAI SDK drop-in compatibility
Standard latency logs
60 req/min rate limit

Every capability, by tier.

Capability	Developer	Production	Enterprise
OpenAI-compatible gateway
Multi-provider routing
Signed compute passes
Public pass explorer
Prompt / response hashing
Solana proof anchoring
Provider reputation ledger
Agent spending policies
Staking / slashing layer

FAQ

Questions at the gate.

Still unsure? Ask the team →

What counts as a request?: One sealed completion through the gate regardless of token count is one request. Streaming responses count once. Embeddings and verification calls are free.
Do you mark up provider token costs?: No. You pay providers their list rate, passed through at cost on the pass. Turnstile only charges the plan fee for routing, sealing, and anchoring.
Is Solana anchoring included?: Off-chain signed passes are included on every plan. On-chain Solana memo anchoring is enabled by default on Enterprise and available as an add-on on Production.
What happens if I exceed my plan?: Requests past your monthly allotment are billed per-thousand at the overage rate, or hard-capped if you set a budget. Sub-keys can carry their own ceilings.
Can I bring my own provider keys?: Yes. Attach your own OpenAI, Groq, or Together keys and Turnstile will route through them while still sealing and anchoring every pass.

Start free. Anchor proofs when you’re ready.

Start building Talk to sales