Chrysent AI Gateway

The problem this solves

AI usage spreads quickly once teams move beyond experiments: multiple model providers, internal copilots, custom agents, growing bills, and security teams asking for evidence you do not have.

At scale, cost, quality, and safety become difficult to see from the outside. Without a central control layer, AI operations remain a black box and you cannot answer simple questions: who spent what, on which model, with which data, and what happened next.

What changes when you adopt Gateway

• Cost attributed at the unit that matters

  Per-user, per-team, per-project, per-application attribution across providers, models, and deployment surfaces — surfaced in operational dashboards.

• Budgets and circuit breakers

  Set spend thresholds by tenant or unit; trigger alerts before overruns; automatically stop runaway calls when budgets are exceeded.

• Immutable session ledger

  Every prompt, response, tool call, and policy decision recorded into an append-only ledger with cryptographic chaining and full-text + semantic search.

• Replay for audits and investigations

  Reconstruct a session end-to-end for compliance reviews, incident investigation, or operational debugging.

How Gateway operates

1. 01

   Configure the Gateway endpoint

   Export an environment variable or change existing client or application configuration to use the Gateway endpoint. No sidecar installation, application rewrite, or application code changes are required.

2. 02

   Apply routing and policy

   Route across providers, apply tenant policies, redact sensitive values, and shape traffic without code changes.

3. 03

   Observe and enforce

   Watch cost, latency, and policy decisions in real time; configure alerts, budgets, and automatic enforcement.

4. 04

   Audit

   Search and replay any session from the immutable ledger; export to your SIEM or data warehouse.

Component map

• Gateway control planeRoutes, enforces policy, and aggregates telemetry.
• Session ledgerAppend-only store with cryptographic chaining.
• Policy engineEvaluates routing, redaction, and budget rules per request.
• Cost & usage dashboardsPer-tenant dashboards, alerts, and budget controls.

For the full platform architecture, see /platform.

Who uses it and how

• AI platform engineering

  Scenario: You expose models to product teams and need central rate limiting, fallback, and audit.

  Gain: A single ingress point with policy-as-code, vendor fallback, and a session ledger you can search.

• FinOps / Tech leadership

  Scenario: You need to attribute AI spend to teams and projects and stop overruns before month-end.

  Gain: Real-time cost attribution, budgets, and circuit breakers wired into the request path.

• Compliance / Risk

  Scenario: You need immutable evidence of every AI interaction for audit or regulator review.

  Gain: A searchable session ledger with policy decision audit and export.

See the product surface

Supported systems

Chrysent AI Gateway integrates with the providers and frameworks below. Each entry shows the current support status. Documentation for the full integration list is coming soon.

• OpenAI (Chat Completions, Responses, Assistants)Generally available
• Anthropic (Messages API)Generally available
• AWS BedrockTech Preview
• Open Source Inference (vLLM, TGI, llama.cpp)Generally available
• Custom OpenAI-compatible or Anthropic Messages-compatible endpointsGenerally available

Where it runs and how data is handled

Chrysent sells self-hosted software only. Software and retained product data remain in customer-owned or customer-controlled infrastructure. Customer-configured external model providers receive provider-bound request data under the customer's provider agreement. This model supports data sovereignty and residency requirements. Data handling and retention are described on the Data handling & deployment page. Security architecture, encryption, and incident response are documented on Security & trust.

Benchmark and evaluation methodology

The first public benchmark study is in preparation. Until it ships, design partners can request the evaluation methodology and a private run report under NDA via the contact form.

Other Chrysent products

• Chrysent Touchstone

  Private beta

  Chrysent Touchstone focuses on protecting AI systems from malicious or unsafe interactions.

• Chrysent Bastion

  Private beta

  Chrysent Bastion protects enterprise systems from unsafe AI generated actions.

Common questions

• Does Gateway add latency?

  Gateway is an AI gateway endpoint, not an application sidecar. Median added latency is in the low single-digit milliseconds for typical traffic; design partners can confirm with their own measurements.

• How is spend attributed?

  By tenant, team, project, application, and any custom label you attach to a request. Attribution flows through to dashboards and export.

• How is Gateway deployed?

  Gateway is self-hosted in customer-owned or customer-controlled infrastructure. The control plane, ledger, and retained product data remain in your environment. Customer-configured external model providers receive provider-bound request data under your provider agreement. This model supports data sovereignty and residency requirements.

• Is the session ledger tamper-evident?

  Yes. Each entry is chained cryptographically; tampering is detectable on read.