New- Context
- 1M
- Max out
- 128K
- Input
- $10.00/M
- Output
- $50.00/M
Enterprise AI Gateway
Make Enterprise AI Sovereign.
Secure, compliant, high-concurrency access to every model — every token, every millisecond and every dollar under one sovereign control plane.
One API. Scale to Infinity.
Requests · today↗ live
0
Req / sec
920
Tokens
4.21B
Throughput920
p50
312 ms
Uptime
99.97%
Recent routed callsauto-failover ON
Models
Open model catalog →The best models, the moment they ship.
Call the newest frontier models the day they go live — fast, current and under one API. Don't see the one you need? We'll bring it on.
New
- Context
- 1M
- Max out
- 128K
- Input
- $5.00/M
- Output
- $30.00/M
- Context
- 1M
- Max out
- 66K
- Input
- $1.50/M
- Output
- $9.00/M
- Context
- 1M
- Max out
- 384K
- Input
- $0.44/M
- Output
- $0.87/M
New- Context
- 1M
- Max out
- 1.0M
- Input
- $1.40/M
- Output
- $4.40/M
- Context
- 512K
- Max out
- 128K
- Input
- $0.30/M
- Output
- $1.20/M
New- Context
- 262K
- Max out
- 262K
- Input
- $1.90/M
- Output
- $8.00/M
- Context
- —
- Max out
- —
- Input
- $0.00/M
- Output
- —
Enterprise capabilities
On a foundation of security, stability and high concurrency.
The operational controls enterprises actually run on — tenant governance, spend control, observability and an open tool ecosystem.
Zero-trust security99.9%+ availabilityHigh-concurrency routing
Tenant management
Org / project / role hierarchy with tiered API keys, isolated workspaces and OIDC & SCIM provisioning.
docs · platform manual →Quota management
Per-key, per-project budgets and rate limits with real-time spend tracking — no more black-box group ledger.
docs · quota management →Alerting & monitoring
Live usage dashboards plus spend, error-rate and anomaly alerts routed to the channels your teams already watch.
docs · alert monitoring →Webhooks
Push request, billing and security events straight into your own systems — reconcile, automate and audit in place.
docs · webhooks →IP whitelist
Lock every API key to trusted networks — the first line of defense for credentials that call frontier models.
docs · ip whitelist →Tool integrations
Drop-in support for OpenCode, Openclaw, Claude Code, Codex and Claude Desktop — open by default, no lock-in.
docs · integrations →Security & Compliance
Your data stays yours.
Security and compliance are defaults, not add-ons — the full credentials wall, with a clear status on every line.
Zero-trust & AI WAFLive
AI-specific threat defense, natively integrated.
Immutable audit logLive
Append-only, cryptographically chained.
OIDC & SCIMLive
Enterprise identity & automated provisioning.
Zero data retentionLive
ZDR by default on production deployments.
SOC 2 Type IIEst. Dec 2026
Independent controls audit — certification estimated December 2026.
GDPROn request
DPA & data-subject controls scoped per deployment — whatever your region requires.
7×24 human supportLive
Not a chatbot — a dedicated engineering channel for your team, reachable in minutes, any hour, any timezone.
Dedicated SlackTelegram groupOn-call engineers
From the blog
All posts →Three signals we publish, every week.
Platform updates the moment they ship, new models the day they go live, and a weekly read on the AI-infrastructure layer.
Product Updates
New platform features, the moment they ship.
Model Updates
Every new model, the day it goes live on AF.
AI Infra Weekly
A weekly read on the infrastructure layer.
Get started
Up and running in minutes.
Bring every model, team and workload under one auditable control plane.