Argus Report

Agent Runtimes

Tracking 15 agent runtimes across 4 languages. Click any agent for the full hub.

Filter:
O

OpenClaw

TypeScriptFoundation

The original open-source personal AI agent

162K stars1.52GB RAM5.98s
v3.2.1Peter Steinberger / OpenAI
L

Letta

PythonActive

Stateful agents with long-term memory

14.2K stars~120MB RAM~3s
v0.6.1Letta AI
P

PicoClaw

GoActive

AI agent for edge devices and IoT

12K stars<10MB RAM~1s
v2.0.1Sipeed
A

AgentKit

PythonActive

Modular Python framework for composable agents

8.4K stars~80MB RAM~2s
v1.2.0Coinbase
m

memU

TypeScriptActive

Persistent memory layer for AI agents

6.9K starsvaries RAMn/a
v1.4.0NevaMind
C

Cortex

TypeScriptActive

TypeScript agent with visual workflow builder

5.6K stars~90MB RAM~2s
v2.1.0Cortex AI
N

Nanobot

PythonActive

Python-native AI agent for research and automation

4.8K stars~50MB RAM~1s
v0.9.3HKU
Z

ZeroClaw

RustActive

Extreme performance Rust agent runtime

3.7K stars7.8MB RAM<10ms
v0.4.2Sundai.Club (Harvard/MIT)
S

Swarmlet

PythonBeta

Multi-agent swarm coordination framework

3.2K stars~200MB RAM~4s
v0.8.0SwarmAI Collective
B

Boltz

GoBeta

Lightning-fast Go agent for microservices

2.1K stars~12MB RAM<100ms
v0.5.0Boltz Labs
T

TinyAgent

GoActive

Minimal agent for resource-constrained environments

1.8K stars<5MB RAM<50ms
v1.0.2TinyStack
N

NanoClaw

TypeScriptActive

Enterprise-grade agent with container isolation

1.2K stars~30MB RAM<1s
v1.1.0Qwibit
F

Forge

RustBeta

Rust agent runtime with hot-reload skills

980 stars~18MB RAM<20ms
v0.3.2Forge Collective
I

IronClaw

RustActive

WebAssembly-sandboxed agent runtime

820 stars~15MB RAM<50ms
v0.3.0Near AI
A

Axiom

RustExperimental

Formal-verification-backed agent runtime

560 stars~20MB RAM<30ms
v0.2.0Axiom AI Research