Gate

The API gateway for AI

One secure endpoint between your company
and every AI model.

Mask providers. Protect credentials. Control where data goes. See every request, token and dollar in real time.

Bitly shortened every link behind one domain. Cloudflare put every site behind one edge. Gate does that for tokens — so token maxers and AI builders stop token maxing and start routing every prompt to the cheapest capable model.

Access Now (beta)

OpenAI · Anthropic · Gemini SDKs — one key, no rewrite

Turn any model key into a Gate key

OpenAI, Anthropic, Gemini — or your own self-hosted endpoint. We make one live call to prove the key works, then reserve your Gate key. Nothing is stored.

Two-way AI security in action (simulated)$18.42 spent this minute

uninspected promptallowed through GatePII removed, then passedstopped at Gate

IdentityProviderPolicySensitive dataBudget

Two decisions. One gate.

Route intelligently. Enforce policy before data leaves.

Gate first inspects each request, then chooses the least expensive approved model capable of doing the work.

01 / SMART ROUTER

The cheapest model qualified for the job

Gate scores each prompt’s complexity, applies your cost and latency rules, then picks the cheapest qualified model from your connected providers. The dashboard shows the price of every route and what you saved versus always using a top-tier model.

prompt before decisionselected routesavings vs top tier

02 / POLICY ENGINE

Every prompt follows your organization’s rules

Gate evaluates privacy, compliance and custom policies before a model sees the prompt. Safe traffic passes, sensitive data is cleaned, and forbidden traffic stops.

uninspectedallowedPII removedblocked at Gate

The problem

Your company is already spending on AI.Nobody can say exactly where it went.

$0K

monthly AI spend

spread across nine invoices, none reconciled

models in production

chosen by whoever shipped first, never revisited

providers reachable

some approved by security, some discovered later

Without a gate

App → OpenAI
App → Anthropic
Agent → OpenRouter
Finance → Gemini
Support → Private model

Different credentials, different providers, no common perimeter.

With Gate.dev

Apps / agents / employees

GATE.DEV

Approved AI infrastructure

One gate. One policy layer. One ledger.

Mask

Your infrastructure stays behind the gate.

One Gate credential. Your provider keys, endpoints and private model locations stay behind the gate.

Provider credentials encrypted at rest, never returned to callers
Destinations, regions and fallbacks hidden from the application
Rotate keys or migrate providers without a code change

Guard

Every request gets checked at the gate.

Identity, provider, policy, sensitive data and budget are evaluated before a single token leaves.

Allowlists for providers, models and regions
Sensitive-data detection with redact, reroute or block
Rate limits and hard budgets enforced in-flight

Observe

See your AI infrastructure moving in real time.

Every request, destination, token, dollar and policy decision — visible the moment it happens.

Live traffic, security and money views of the same stream
Per-request receipts kept as audit evidence
Spend attributed to team, project, app and agent

How it works

Three steps. No re-platforming.

Point one URL at the gate

Swap the base URL in whatever SDK you already use — OpenAI, Anthropic or Gemini. Keep the native model names your code already sends; Gate resolves and routes them.

- base_url = "https://api.openai.com/v1"
+ base_url = "https://api.gate.dev/v1"
- api_key  = OPENAI_API_KEY
+ api_key  = GATE_API_KEY

Set the rules once

Who can reach which model, what data may leave, which regions are approved, and where the money stops. Rules apply at the gate, so every app inherits them the moment they ship.

rule "no-pii-offshore" {
  when   prompt.contains(pii)
  then   redact() and route("eu-private")
}
budget "support" { cap 40k/mo -> stop }

Watch the traffic and the money

Live map, cost ledger, treasury projections and savings opportunities — one shared picture for finance, security and engineering, exportable as CSV or a signed PDF.

Requests today42,891

Spend MTD$18,421

Blocked / redacted127

of AI spend recovered in quarter one

of requests logged with full evidence

0ms

median gateway overhead

0 lines

to adopt, and two to leave

Who runs through the gate

Same gate, different reasons.

Each organisation shape uses Gate for a different question — cost, sovereignty, secrecy or speed. Token maxers come for the routing; enterprises come for the audit trail.

Enterprises

Global, many teams, many vendors

Chargeback AI spend per team, project and cost centre
Kill shadow AI: no key works outside the gate
Swap models centrally without touching product code

31% typical spend cut in the first quarter

Governments

Public sector & defence

Residency rules: block requests leaving approved regions
Full audit trail per request for oversight and FOI
Classification-aware redaction before anything leaves

100% of prompts retained for audit

Private companies

Founder-led, IP-sensitive

Redact contracts, salaries and source code in-flight
Hard monthly budgets that stop, not just warn
One invoice instead of nine provider accounts

0 unreviewed vendors reachable

Financial services

Banks, insurers, funds

Immutable request ledger for regulators and internal audit
PII and account-number detection on every prompt
Alerting on spend spikes and model drift

7 yrs exportable retention window

Healthcare & research

Providers, payors, labs

Route sensitive workloads to on-prem models by policy
Per-study cost tracking for grants and trials
Prove what data did — and did not — leave

1 gate for every clinical AI workload

Startups & AI-native

Shipping fast on thin margins

Opportunities engine suggests cheaper model swaps
Per-customer cost so you can price with confidence
Instant failover when a provider degrades

4.2x cost-per-request improvement found

Token maxers & AI builders

Prompt engineers, agents and automation teams

Route every prompt to the cheapest capable model automatically
See token count and cost before the request is sent
Cap per-key, per-app and per-project spend so token maxing stops

40% typical token-bill reduction with smart routing

Gate vs. them

Prompt-security tools watch employees.Gate governs the API traffic your company actually runs on.

Wald, Obsidian and Prompt Security guard the laptop and the browser. LLM gateways route tokens but defend nothing. Gate is the only layer that masks outbound data, blocks what the model sends back, and prices every token — in one perimeter, with no agent to install.

They watch people. We govern machines.

Endpoint DLP inspects what an employee types into a chatbot. Gate sits on the API path, where your applications, agents and jobs generate the overwhelming majority of tokens.

Two-way, not one-way.

Everyone filters the prompt going out. Gate also treats the model's answer as untrusted input — injected instructions, rogue tool calls and exfiltration attempts are blocked on the return leg.

Security and FinOps in one ledger.

A blocked request and an expensive request show up in the same receipt. Security vendors can't price your traffic; gateways can't defend it.

Infrastructure, so it can't be bypassed.

No agent to install, no browser extension to disable. If the key doesn't route through Gate, the call doesn't happen at all.

Capability	Gate.dev	Employee AI DLPWald · Obsidian · Prompt Security	LLM gatewaysLiteLLM · OpenRouter · Portkey
One key in front of every provider OpenAI, Anthropic, Gemini, Bedrock, Azure and self-hosted behind one endpoint.
Outbound PII / PCI masking before egress DLP tools do this for humans. Gate does it for every API call.
Inbound defence against the model itself Poisoned answers, injected tool calls and exfil attempts are stopped on the way back.
Cost ledger, budgets and chargeback Per team, per key, per department — with hard ceilings, not warnings.
Routing, failover and model allow-lists Swap models centrally without shipping product code.
Immutable per-request receipts Reconstruct any call years later: model, policy trace, tokens, cost.
Covers agents, pipelines and cron jobs Non-human traffic is where AI volume actually lives.
Provider keys sealed in a vault Envelope-encrypted, rotated, never handed to application code.
Works with zero code change Change the base URL and the key. That's the migration.
Endpoint agent for consumer chatbots Their home turf — deliberately not ours. Gate complements it.

API perimeter (infrastructure)

Gate.dev

One key in front of every model API. Masks data outbound, guards answers inbound, meters and routes every token.

Deployment:: Nothing to install — swap the base URL and key
Secures:: Machine traffic: apps, agents, pipelines, jobs
Wins:: Machine-to-model traffic: your apps, agents, pipelines and jobs.
Gap:: Not an endpoint agent for employees pasting into consumer chatbots.

Endpoint AI DLP

Wald.ai

Local agent sanitises prompts and file uploads before they reach ChatGPT, Claude or Copilot.

Deployment:: Agent installed on managed endpoints
Secures:: Humans typing into consumer AI apps
Wins:: Employee shadow-AI on managed laptops.
Gap:: No control over API traffic, no routing, no cost ledger.

Browser / SaaS posture

Obsidian Prompt Security

Extends SaaS security posture management to GenAI apps used in the browser.

Deployment:: Browser extension / SaaS-connected posture
Secures:: Browser-based GenAI SaaS usage
Wins:: Discovering which AI SaaS employees signed up for.
Gap:: Blind to server-side calls; nothing about spend or model choice.

Prompt firewall (human + some API)

Prompt Security

Inspects prompts and responses for injection, leakage and policy violations.

Deployment:: Inline proxy or browser extension
Secures:: Prompt and response content
Wins:: Prompt-level threat inspection.
Gap:: Security only — no unified key, no failover, no FinOps.

Read the full breakdown

Gate vs. Cloudflare AI Gateway

Same proxy idea.Different layer, different buyer.

Cloudflare AI Gateway is an edge control plane for teams already on Cloudflare. Gate is a standalone token-security and routing layer: security on by default, budgets and chargeback built in, and nothing to migrate — it works from any cloud with any provider key.

Capability	Gate.dev	Cloudflare AI Gateway
Deployment model	Standalone managed service — works from any cloud, VPC or laptop	Runs on Cloudflare's edge network
OpenAI-compatible drop-in endpoint	Swap base URL and key — no Cloudflare account or DNS move needed	Swap base URL and key
Outbound PII / PCI masking	Deterministic tokenisation of emails, phones, cards, SSNs, secrets before egress	Via Firewall for AI add-on
Prompt-injection & jailbreak blocking	Built in, on by default, no WAF rules to write	Via Firewall for AI add-on
Signed provider-key vault	Envelope-encrypted, password-gated reveal, never exposed to app code	BYOK support
Smart routing to the cheapest capable model	Complexity tiers + custom rules by cost, latency, context and capability	Fallbacks, retries and weighted routing
Per-key budgets with hard ceilings	Monthly limits enforced before the request leaves, alerts in-console	Not offered — analytics only
Department / employee attribution & chargeback	Every token tagged to a team, project and person	Not offered
Request caching	Yes, keyed per request	Yes — strongest caching story in the category
Self-hosted / OpenAI-compatible models	First-class: Ollama, vLLM, any OpenAI-compatible endpoint	Worker AI + major hosted providers
Immutable per-request receipts & exports	CSV / PDF reports of spend, ledger and requests, filtered	Usage analytics & logs
Edge network, DDoS absorption & Workers runtime	Not our layer — we sit beside any edge, not in place of one	Deeply integrated with the Cloudflare platform

Choose Gate.dev when

You want prompt security, smart routing, budgets and department chargeback without moving DNS or adopting a platform — a drop-in key that works with any provider, from any cloud.

Choose Cloudflare when

Your infrastructure already lives on Cloudflare and you mainly need caching, rate limiting and logs at the edge. Gate complements that edge — it is the token-security and cost layer on top.

Based on Cloudflare's published AI Gateway documentation, September 2026. Features evolve — verify before purchase.

Questions

The ones security asks first.

Get started

Put a gate in front of it.

One base URL change and every call is metered, policed and receipted. Walk the demo organisation with live data, or get the onboarding guide in your inbox.

Access Now (beta) Read the docs

No spam. Invite-only beta; access codes are sent by email.

5-minute install

Swap the base URL

Keys stay yours

Vaulted, never logged

Audit-ready

Signed per-call receipts

One secure endpoint between your company and every AI model.