agentboards.org
Compare/agentUniverse vs Legion

agentUniversevsLegion

Generated from the two spec rows. Green marks the better value where a spec has a clear direction. Everything else is just different.

agentUniverse
Ant Group · Agent framework
OSSMCP
Panel
6.4
4 spec wins
Reliability
6.2
Usefulness
6.0
Cost
7.0
Longevity
6.5

“Two of its patterns are acronyms, which is how you know the framework came out of a bank.”

Legion
Software Mansion · Agent framework
OSS
Panel
6.5
2 spec wins
Reliability
6.0
Usefulness
6.2
Cost
8.0
Longevity
5.8

“It writes Lua to get work done in Elixir, which is two functional languages agreeing that a third one should take the blame.”

Spec by spec

SpecagentUniverseLegion
Architecture
CategoryAgent frameworkAgent framework
Runslocallocal
Platformsmacos, linux, windowsmacos, linux
Context windownot documentednot documented
Protocols
MCP clientYesNo
MCP serverYesNo
Capabilities
Runs terminal commandsNoNo
Multi-file editsNoYes
Git operationsNoNo
Browser controlNoNo
Sandboxed executionNoYesLegion runs generated Lua or Elixir in its own sandbox; the README does not describe a container runtime.
Multi-agent orchestrationYesNo
Headless / CI modeNoNo
Models
BackboneQwen, DeepSeek, OpenAI, Claude, Gemini, Llama, Kimimultiple providers
Bring your own modelYesYes
Local modelsNoNo
Cost
Pricing modelbyokbyok
Starts at$0/mo$0/mo
Free tierYesYes
Bring your own keyYesYes
Openness
Open sourceYesYes
LicenseApache-2.0MIT
GitHub stars2,374220

Which one would each critic pick

CriticagentUniverseLegionPick
El Juez——not enough reviews
El Amigo6.56.8Legion — Pick Legion if you write Elixir and want the agent inside the application rather than beside it; pick a framework in Python if the rest of your stack already lives there.
El Crítico5.86.0Legion — The generated snippet runs in the project's own sandbox, and the README does not describe a container runtime, so the isolation boundary is whatever the host process could already reach.
El Profesor6.57.8Legion — One evaluation can filter, branch and loop, so a sequence that costs a tool-calling agent a model round trip per step collapses into a single call. The claim is structural and checkable.
La Inversora6.06.3Legion — 166 stars and a consultancy's name on it. Software Mansion builds open libraries to sell engineering time, which funds this exactly as long as the pitch still works.
La Jefa6.05.5agentUniverse — No seat cost for sixty engineers and no vendor either, so support is a public issue tracker and the meter is whichever model vendor the configuration points at.
El Hacker7.86.8agentUniverse — Apache-2.0, pip install agentUniverse, and it speaks MCP in both directions, so my servers plug in and the whole thing can be published as one.

Picks are derived from each critic's own scores. Humans vote on matchups on the duels page.

Want a third column? The compare tool handles any two agents. Three-way comparisons are on the roadmap once the spec rows are all verified.