#9696th of 118 in Agent frameworks and SDKs
oasis
LLM-agent social media simulator for Twitter and Reddit-style platforms
- Stars
- 5.4k
- License
- Apache-2.0
- Last commit
- Nov 2026
- Last release
- Nov 2026
- Language
- Python
4,321 on 21 Apr 20265,414 on 10 Nov 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
OASIS is a Python library that simulates Twitter and Reddit-like platforms populated by LLM-driven agents, storing state in a database file. Agents choose from 23 actions (post, comment, follow, repost, search and others) and see content ranked by interest-based or hot-score recommenders. It targets research on information spread, polarization and herd behavior, with a stated scale of up to one million agents.
Who it is for: Researchers simulating social media dynamics with LLM agents
Strengths
- 23 agent actions, including follow, repost, quote, group chat and interview
- Interest-based and hot-score recommendation algorithms built into the simulated platform
- Per-agent model, tools and prompt customization, with a PettingZoo-style env interface
- Installs from PyPI as camel-oasis; paper, dataset and docs site published
Weaknesses
- Token cost grows with agents and steps: 100 agents, one step used about 336k input tokens
- Quick start only shows OpenAI; other providers go through CAMEL's model factory
- Multi-modal platform support is still an open issue, not shipped
- Last tagged release is December 2025, well behind the latest commits
What it needs
- no GPU
- Needs camel-ai, LLM API key (OpenAI in quick start)
- Models: OpenAI GPT-4o-mini, Qwen (via CAMEL), OpenAI embeddings for Twhin-Bert recommender
Quick answers
Is oasis open source?
Yes. Its code is on GitHub under the Apache-2.0 license.
Does oasis need a GPU?
No, it runs without one.
Also in Agent frameworks and SDKs
See all 118| Rank | Project | Score |
|---|---|---|
| 1 | ppt-masterAgent skill that turns documents into natively editable PowerPoint decks | 95 out of 100 |
| 2 | openai-agents-pythonPython SDK for multi-agent workflows with handoffs, guardrails and tracing | 91 out of 100 |
| 3 | mastraTypeScript framework for building AI agents, workflows and MCP servers | 91 out of 100 |
| #4 | browser-usePython library that lets an LLM agent control a web browser | 90 out of 100 |
| #5 | copilotkitSDK for adding agent chat, generative UI and shared state to apps | 90 out of 100 |