Now in Early Access

We believe your AI should answer to you.

Not to a vendor. Not to a pricing page. Not to someone else's terms of service. Everyone else sells you a rented brain. We sell you the mine.

94%
Local Hit Rate
<200ms
TTFT
$0
Local Inference
$0.40
Per 1M Fallback

The Golden Circle

Start with why. Everything else follows.

WHY

AI is too important to rent

Your models, your data, your decisions. When the most transformative technology in history is controlled by someone else, you've already lost the plot.

HOW

One stack, your machines, no compromises

Riven gives you a unified AI infrastructure layer that runs on hardware you own, with cloud fallback when you need it. No vendor lock-in, ever.

WHAT

Infrastructure that stays on your side

A complete AI stack — gateway, agents, automations, knowledge base, and browser — that lives on your side of the firewall.

Deploy your way

Three deployment modes. One unified experience. Choose the level of control that fits your organization.

Fastest Start

Riven Cloud

We host it. You use it.

Zero-config hosted Riven. Get an API key and start building in minutes. Your data is encrypted and isolated.

  • Managed infrastructure
  • Auto-scaling
  • 99.9% uptime SLA
  • Global edge network
Most Popular

Riven On-Instance

Your VM. Your rules.

Deploy Riven to your own virtual machines or bare metal. Full control, full sovereignty, full stack.

  • Docker / Docker Compose
  • Single-node or clustered
  • Your cloud, your data center
  • Air-gapped compatible
Enterprise

Riven Managed

Your racks. Our ops.

We manage Riven on your infrastructure. Enterprise SLAs, dedicated support, custom integrations.

  • Dedicated support engineer
  • Custom SLAs
  • On-prem deployment
  • Compliance assistance

The full stack

Six components. One infrastructure. Everything you need to run AI on your terms.

Gateway

OpenAI-compatible API gateway. Route to local models first, cloud fallback second. One endpoint, every model.

Agents

9 autonomous personas that think, plan, and execute. Task decomposition, tool use, and memory built in.

Automations

n8n-powered workflow engine. Connect AI to your business logic with drag-and-drop automation pipelines.

Console

Web dashboard for everything. Monitor usage, manage models, configure agents, review logs. All in one pane.

Relay

AI-native browser with baked-in Riven integration. Sidebar chat, voice input, document RAG, smart new tab.

RAG

Qdrant-powered knowledge base. Upload documents, chunk automatically, retrieve with semantic precision.

Architecture

Five layers. Local-first. Cloud when you want it.

USER LAYER
Relay BrowserConsole DashboardAPI Clients
GATEWAY LAYER
OpenAI-compatible APIRequest RoutingAuth & Rate Limits
AGENT LAYER
9 PersonasTask PlanningTool ExecutionMemory
AI BACKEND
Local Models (Ollama)RAG (Qdrant)Automations (n8n)
CLOUD FALLBACK
OpenAIAnthropicGoogleOn-demand only

How we compare

The only platform that gives you the full stack — gateway, agents, browser, and automations — in one self-hosted package.

FeatureRivenAnythingLLMDifyOnyxOpenWebUI
Self-hosted
OpenAI-compatible API
Autonomous Agents
Workflow Automations
Built-in RAG
Web Dashboard
Custom Browser
Cloud Fallback
Air-gap Ready
Multi-model Routing
SSO / SAML
RBAC
Audit Logging
Voice Input
Enterprise Support

Simple, transparent pricing

Start free. Scale when you're ready. No surprises.

Free

$0forever

Try Riven with zero commitment.

50 queries/day
  • Riven Gateway access
  • 1 agent persona
  • Community support
  • Relay browser
Start Free

Starter

$29/mo

For individuals and small teams getting serious.

3,000 queries/mo
  • Full Gateway access
  • 3 agent personas
  • Basic automations
  • Email support
  • Console dashboard
Get Started

Pro

$99/mo

For teams that need unlimited power.

Unlimited queries
  • Everything in Starter
  • All 9 agent personas
  • Advanced automations
  • RAG knowledge base
  • Priority support
  • SSO / SAML
Go Pro

Enterprise

Custom

For organizations with custom requirements.

Unlimited + SLA
  • Everything in Pro
  • Dedicated support engineer
  • Custom integrations
  • On-prem deployment
  • Compliance assistance
  • Audit logging
Contact Sales
Manifesto

Something went wrong with AI.

It was supposed to be the great equalizer. Instead, it became the great extractor. We wrote a manifesto about why we're building Riven — and what we think the future of AI should look like.