Features Pricing About Get access →
PRICING

Simple. Transparent. Self-hosted.

You pay for the software. Cloud model calls pass through at cost — no markup, no surprise invoices.

SELF-HOSTED

Open Core

$0
forever, self-hosted
  • Full OpenAI-compatible gateway
  • Unlimited local GPU inference
  • Ollama + any local backend
  • Basic routing (explicit model IDs)
  • In-memory session context
  • REST API + streaming SSE
  • Community support
Read the docs →
RECOMMENDED

Pro

$49
per month · self-hosted + managed features
  • Everything in Open Core
  • Auto-routing (MoE classifier)
  • Persistent semantic memory (pgvector)
  • LSR reasoning pod
  • MCP tool server
  • Multi-tenant API key management
  • Audit log & cost attribution
  • Policy-based routing rules
  • Email support
Get Pro access
ENTERPRISE

Enterprise

Custom
volume discounts · white-glove onboarding
  • Everything in Pro
  • Air-gapped / VPC deployment
  • SSO / SAML / LDAP
  • Custom model fine-tuning integration
  • Dedicated routing cluster
  • SLA + uptime guarantee
  • Onboarding engineering support
  • Dedicated Slack channel
Talk to us
CLOUD PASS-THROUGH

You pay your cloud providers directly — no markup

When you route requests to Anthropic, OpenAI, Gemini, or any other cloud backend, Dream-Weaver passes through your API key. We never see your cloud bills, never mark up usage. You pay vendors directly at their published rates.

Questions? Just ask.

Early access pricing is flexible. If you're running local GPU hardware and want to try Pro, reach out — we'll work something out.

Get in touch