Hybrid AI Orchestration Platform
Hybrid AI orchestration platform — run any AI model (local, cloud, or self-hosted) from a single workspace. BYOK to any provider, build agents and apps, no per-seat fees.
Try it →
|
Pricing
|
Compare
Features
BYOK any cloud API, run open weights on your laptop, or deploy a dedicated GPU endpoint — all from one workspace.
Fine-tune Llama, Mistral, or Qwen with LoRA on your data. Quantise for cheap inference. Hot-swap adapters at runtime.
Persistent agents that schedule, plan, use tools, share memory, and run 24/7 — without you watching.
Ship a tool, hit Deploy, get a public URL. Database, secrets, cron, logs all included. Replaces team chat, CRM, and dashboard tools.
Auto-extracted entities, knowledge graphs, custom retrieval pipelines, and reranked RAG — all configurable in osStudio.
Web IDE, GitHub import, PR reviews, and Maestro chat across your repos — without leaving the workspace.
Native realtime channels, change-data capture, hosted OAuth, and a fleet of MCP servers — all built into the workspace.
AI agents that browse logged-in sites, scrape pages on a schedule, and ingest from 2,000+ news sources.
Edit every prompt, pipeline, agent, retrieval stage, routing rule, and tool as a versioned config. Fork, branch, share.
Run models on-device, keep private channels off the cloud, and work offline. Your data, your machine.
WorkOS SSO, audit logging, RBAC, anomaly detection, secret rotation, GDPR data erasure, multi-region residency, and BYO-cloud.
Pay only for what you use. 250K open datasets indexed. Page builder, website builder, and public KB hosting included.
How osFoundry compares to the alternatives
AI glossary — terms with osFoundry’s angle on each
Frequently asked questions
What is osFoundry?
osFoundry is a hybrid AI orchestration platform that lets you run any AI model — local, cloud API, or self-hosted GPU — from a single workspace. It bundles a chat agent (Maestro), a visual orchestration config editor (osStudio), data-backed mini-apps (Room Apps), code repos, knowledge bases, and a connector library, all metered as pure usage with no per-seat fees.
Is osFoundry open source?
osFoundry is source-available — the desktop app, the Room App SDKs, and example apps are public, while the managed cloud platform is operated as a service. You can self-host the entire platform in your own AWS, GCP, or Azure account under the BYO Cloud plan.
Is osFoundry free?
The platform has no monthly fee. You pay only for what you use: tokens through your own provider keys (BYOK), compute time for container apps, and storage. New workspaces receive free credits to evaluate, and any local-only workflow running on your own machine costs nothing.
How much does osFoundry cost?
Pricing is fully usage-based. There is a $10 minimum credit top-up; from there you pay per token (passed through from your provider at a small markup when using the managed proxy), per compute-minute for container apps, and per GB-day for storage. No seat fees, no minimums, no annual contracts.
Does osFoundry support local AI models?
Yes. osFoundry runs open-weight models locally via llama.cpp on macOS, Windows, and Linux. The desktop app detects your hardware, picks an appropriate quantization, and exposes the local model to Maestro, agents, and Room Apps through the same interface as cloud models. Inference never leaves your device.
Can I bring my own API key (BYOK) to osFoundry?
Yes — BYOK is the default for every cloud provider. Paste an OpenAI, Anthropic, Google, Mistral, xAI, Groq, Together, Fireworks, or DeepSeek key into the workspace and that model becomes available to every agent, Room App, and pipeline. Keys are encrypted in the workspace and never logged.
How is osFoundry different from ChatGPT or Claude?
ChatGPT and Claude are hosted chat tools — one model, one chat surface, per-seat invoice. osFoundry is the orchestration layer around any model: schedule agents on a cron, build internal apps with persistent databases, host open-weight models locally, route the same prompt across multiple providers, and version-control every prompt, pipeline, and tool.
How does osFoundry compare to LangChain?
LangChain is a Python framework that you assemble and operate yourself. osFoundry is a managed platform with the orchestration prebuilt — a visual osStudio editor, scheduled agent runs, BYOK to every provider, an included database and storage per app, and pure-usage billing. You can still write custom JavaScript plugins in osStudio for the pieces a framework would handle.
Does osFoundry work offline?
Yes for local-model workflows. The desktop app runs llama.cpp locally and stores notes, chats, and indexes on-device. Cloud features (shared databases for Room Apps, BYOK to cloud providers) require connectivity, but a local-first private mode keeps every byte of data on your machine.
Where is my data stored?
Data location depends on plan. The managed cloud stores workspaces in the US (default) or Japan (selectable per workspace). The BYO Cloud plan runs the platform inside your own AWS, GCP, or Azure account so data never leaves your perimeter. The desktop app stores local data in encrypted SQLite on your machine.
Can I self-host osFoundry?
Yes, via the BYO Cloud plan. The platform deploys inside your own AWS, GCP, or Azure account with your own KMS keys. You keep full data sovereignty and infrastructure ownership; osFoundry maintains the runtime, control plane, and SDK updates.
Does osFoundry support GDPR, SOC 2, or data residency requirements?
Yes. Workspaces support per-region storage pinning (US, EU, JP), GDPR-aligned data deletion endpoints, tamper-evident audit logs, real-time SIEM webhook streaming, and WorkOS-backed SSO with any IdP. BYO Cloud customers can pin everything inside their existing compliance perimeter.