Developer
For individual builders, staging environments, and prototype validation.
- 50,000 requests / month
- 0-token prefill veto
- .sent attestation hashes
- Public notary lookups
Cryptographic execution control for high-stakes AI. SentinelGate halts adversarial injections and tool hijacks at Token 0—evaluating prompt kinematics in-memory with sub-millisecond overhead, before execution, and before token spend. Every action is sealed with tamper-evident Ed25519 evidence bags engineered for **ISO 20022 financial rails**.
Run a public kinematic preview — no API key required. Emerald = permit. Coral = Phase III topological collapse.
{
"status": "resting",
"note": "Awaiting audit — Run Audit or select a preset"
}
Provisioned automatically via billing webhooks — keys minted as sk-sent-XXXX into the Sentinel ledger.
For live production systems requiring high throughput and strict veto enforcement.
Dedicated clusters, custom kinematic threshold tuning, and financial rail SLAs.
XDP_DROP sub-50µs socket severing)
or email directly: [email protected]
Point any OpenAI-compatible client at SentinelGate — swap the base URL and keep your existing SDK.
from openai import OpenAI
client = OpenAI(
api_key="sk-sent-XXXX",
base_url="https://api.sentinelgate.dev/v1",
)
completion = client.chat.completions.create(
model="meta-llama/Llama-3.3-70B-Instruct",
messages=[{"role": "user", "content": "Execute financial database reconciliation."}],
)
Differentiation, financial rails, integration, retention, veto mechanics, and key operations.
SplmtryData) envelopes.
base_url to https://api.sentinelgate.dev/v1 and pass your sk-sent-... API key. Prompts, schemas, streaming, and tool calls work out of the box with zero code refactoring.
kinematic_security_veto, and drops the socket before upstream tokens are consumed. The client receives an Ed25519-signed evidence bag detailing the thermodynamic state coordinates and attestation hash for audit logging.