Autogen

App in the BluixApps catalog

What it is

AutoGen is Microsoft Research's framework for multi-agent LLM workflows — define agents with roles, let them converse to solve tasks. AutoGen Studio is the visual builder on top: design agent teams, orchestrate conversations, run experiments in a browser.

What it's for

  • Multi-agent prototyping — visualize agents collaborating
  • Research experimentation — test agent architectures
  • AI workflow design — manager + worker agents
  • Code-gen workflows — coder + reviewer + tester agents
  • Educational demos — visualize agent reasoning

Who it's for

  • AI researchers experimenting with multi-agent patterns
  • AI engineers prototyping team workflows before code
  • Educators demonstrating multi-agent concepts visually
  • Power users building autonomous agent crews
  • Microsoft Research community + AutoGen library users

Why teams pick AutoGen over alternatives

  • MIT (AutoGen) + Microsoft-backed — sustained development
  • Research-grade — used in published AI research
  • Multi-agent native — conversation patterns built in
  • Tool integration — agents can call functions
  • Studio + library — visual + code paths
  • Active development — major releases tracked

Integrations

  • LLM providers — OpenAI, Anthropic, Azure OpenAI, Ollama, any OpenAI-compatible
  • Code execution — sandboxed Python via Docker
  • Web browsing — agent can use browser tools
  • Custom tools — Python function registration
  • Memory — agent state persistence
  • Multi-modal — vision via GPT-4V, Claude vision
  • Skills marketplace — community-built agent capabilities

Notable users & community

  • 36k+ GitHub stars (AutoGen library)
  • Backed by Microsoft Research
  • Active Discord + GitHub Discussions
  • Used in AI research papers + corporate experiments
  • Standard multi-agent framework in OSS

Tips & operations

  • Research-grade software — breaking changes possible between versions
  • API key required — configure LLM provider in Settings before use
  • Resource-intensive — agent conversations consume tokens fast
  • Production = library not Studio — Studio is for prototyping; production uses code
  • Sandbox code execution carefully — agents can run arbitrary code
  • Conversation logs — review to debug agent behavior

What we ship in BluixApps

  • Docker image: ghcr.io/microsoft/autogen-studio:latest
  • Port 8082 exposed (mapped from 8081)
  • Persistent volume: /opt/autogen for workflows + chat history
  • No built-in auth — protect with reverse proxy in production
  • HTTPS via Let's Encrypt reverse proxy
  • Pre-configured to detect Ollama on same VPS
  • Backup hook covers /opt/autogen data
Read this app's deep dive on bluix.app ↗

Get this app — pick a BluixApps plan

Same catalog. Scaling tenant isolation, white-label and support tier.

TierTenantsCatalogSupportWhite-labelMonthly
Stacks119 curated stacksStandard$19/moDetailDeploy
Starter10Full catalogStandard+$15–25/mo$49/moDetailDeploy
Pro25Full catalogPriority bugfix+$15–25/mo$149/moDetailDeploy
Growth100Full catalogPriority bugfix+$15–25/mo$349/moDetailDeploy
Scale500Full catalog7-day window+$15–25/mo$799/moDetailDeploy
EnterpriseUnlimitedFull catalogPriority 7-dayBundled$1,499/moDetailDeploy

Powered by WHMCompleteSolution