Why AgentsForge
Build experts, wire documents, and chat — without leaking your data.
Privacy-first
Self-host or managed. Your data and documents stay under your control.
Document ingestion
Upload files and retrieve grounded answers with source snippets.
Streaming chat
Fast, token-efficient streaming with multi-provider support.
Citations & source snippets
Trust every answer. See the exact passages used to respond.
Multi-provider ready
Use Ollama, swap models anytime.
Flexible deployment
Run AgentsForge self-hosted in your stack, or on our managed beta cloud.
How it works
From zero to a private expert in minutes.
1. Create
Spin up your expert in minutes.
2. Ingest
Drag & drop docs, we chunk and index automatically.
3. Chat
Ask, cite, and ground answers instantly
We're in beta — join today
AgentsForge is evolving fast. Join our beta, test features, and help shape the roadmap before pricing launches.
Feedback first, plans later.
FAQ
Quick answers to common questions.
Is my data used to train your models?▼
No. Your prompts, documents, and outputs are never used to train shared models. Self-host or use our managed cloud—your data stays yours.
Can I self-host?▼
Yes. AgentsForge is designed to run self-hosted or managed. The product experience is consistent across both.
Which models are supported?▼
llama3.1:8b, mistral:7b, deepseek-r1:8b, and more. We're provider-agnostic; add your own.
How are documents processed?▼
Documents are chunked and embedded (pgvector). Answers include citations with source snippets.
Do you store conversations?▼
You control retention. For auditability, you can keep logs with snippet references.
Is there an on-prem option?▼
Yes. Talk to us for enterprise deployments and SSO.
How can I join the beta?▼
Click 'Join Beta' on this page — it's free during beta.
Ready to build your first expert?
Spin one up in minutes. Keep your data private.
Built with privacy in mind
