camel-ai.org is working on finding the scaling laws of agents. The first and the best multi-agent framework. Discord: discord.camel-ai.org. Product @Eigent_AI
- Tweets 1,461
- Following 95
- Followers 8,592
- Likes 1,581
Pinned Tweet
Open AI and Anthropic just release GPT-5.3-Codex and Opus 4.6 model, terminal capability is now on top of their list evaluating modal capability. But terminal training hits a wall fast: there aren’t enough high-quality environments.
In SETA, we just shipped 1,376 validated terminal environments across: SE • sysadmin • security • debugging • networking • DevOps
Compatible with Terminal Bench & Harbor. @Mike_A_Merrill @alexgshaw
And we’re scaling fast 👀
Find it in: github.com/camel-ai/seta-env or search for seta-env in on harbor registry
4
6
39
6,007
Running OWL is now super easy 🦉
We’ve added a simple UI
(just click, run, and watch your agents work)
Simple, clean, and fast.
11
38
276
16,945
What’s inside the best open-source general AI agent?
Let’s talk about OWL by CAMEL-AI ↓
5
34
247
122,787
We tried OWL with 𝐆𝐞𝐦𝐢𝐧𝐢 𝟐.𝟓 𝐏𝐫𝐨
(the results were seriously impressive)
We gave the agent a task:
"Research Gemini 2.5 Pro, get its benchmark scores, write Python code to plot them, and run it."
It did everything on its own:
✔️ Collected benchmark data
✔️ Wrote clean python code
✔️ Saved both the chart + code locally
No manual steps. Just full-on autonomous execution.
This is how agent workflows should feel in 2025.
You can try it, check the code, or run your own workflows: github.com/camel-ai/owl
3
33
243
20,253
Introducing 🏝️ OASIS: Open Agent Social Interaction Simulations with One Million Agents
🏝️ OASIS is a scalable, open-source social media simulator that integrates large language models with rule-based agents to realistically mimic the behaviour of up to one million users on platforms like Twitter and Reddit.
Here are 4 of its key features:
- 📈 Scalability: OASIS supports simulations of up to one million agents, enabling studies of social media dynamics at a scale comparable to real-world platforms.
- 🔄 Dynamic Environments: Adapts to real-time changes in social networks and content, mirroring the fluid dynamics of platforms like Twitter and Reddit for authentic simulation experiences.
- 👍 Diverse Action Spaces: Agents can perform 21 actions, such as following, commenting, and reposting, allowing for rich, multi-faceted interactions.
- 🔥 Integrated Recommendation Systems: Features interest-based and hot-score-based recommendation algorithms, simulating how users discover content and interact within social media platforms.
💻 Check out the repository: github.com/camel-ai/oasis
✍️ Read the paper: arxiv.org/abs/2411.11581
🌐 Find out more via the project page: oasis.camel-ai.org/
🐫 Join our community: discord.camel-ai.org
12
38
227
29,294
𝐒𝐜𝐫𝐚𝐩𝐢𝐧𝐠 𝐭𝐡𝐞 𝐢𝐧𝐭𝐞𝐫𝐧𝐞𝐭 𝐣𝐮𝐬𝐭 𝐠𝐨𝐭 𝐚 𝐥𝐨𝐭 𝐦𝐨𝐫𝐞 𝐢𝐧𝐭𝐞𝐫𝐞𝐬𝐭𝐢𝐧𝐠.
@firecrawl MCP X OWL is:
→ Fast
→ Lightweight
→ No browser overhead
→ Fully autonomous
→ Structured output in seconds
Time to build!
3
33
226
14,981
Introducing 🦀 CRAB: Cross-environment Agent Benchmark for Multimodal Language Model Agents
🦀 CRAB provides an end-to-end and easy-to-use framework to build multimodal agents, operate environments, and create benchmarks to evaluate them, featuring three key components:
- 🔀 Cross-environment support - agents can operate tasks in 📱 Android and 💻 Ubuntu.
- 🕸️ Graph evaluator - provides a fine-grain evaluation metric for agents.
- 🤖 Task generation - composes subtasks to automatically generate tasks.
By connecting all devices to agents, 🦀CRAB unlocks greater capabilities for human-like tasks than ever before.
Use 🦀 CRAB to benchmark your multimodal agents!
- 👨💻 Check out the repository: github.com/camel-ai/crab
- 📝 Read the paper: arxiv.org/abs/2407.01511
- 🌐 Find out more via the project page: crab.camel-ai.org/
- 🐫 Join our community: discord.gg/8zQTBNqf97
7
45
204
72,673
📢 We've just introduced a knowledge graph agent into the 🐫 CAMEL framework.
It works by first processing the given text with @UnstructuredIO. Then, this agent extracts relationships between entities from the processed content, structuring data more efficiently so it can easily be stored in a knowledge graph databases such as @neo4j.
🔗 Try it out: colab.research.google.com/dr…
Thanks to our contributor @ttokzzzzz for this. 🤝 Explore more here: github.com/camel-ai/camel/pu…
3
40
185
121,657
𝐎𝐖𝐋 × 𝐐𝐰𝐞𝐧𝟑 × 𝐌𝐂𝐏 𝐢𝐬 𝐦𝐚𝐠𝐢𝐜𝐚𝐥 ✨
We demoed agents fetching the Qwen3 GitHub repo, summarizing its intro, and auto-generating HTML documentation site with code examples
then deploying it via MCP and opening it in the browser seamlessly.
Experience the magic.
@Alibaba_Qwen x @CamelAIOrg
5
26
174
15,541
What if your tweets entered a parallel universe where AI agents replied, reposted, and battled for clout?
Meet Matrix — the social simulation engine for social media.
➕ Add any account
📝 Drop a post
🧠 Let agents engage
Try posting and see how the agents reacts, have fun! 👀
→ matrix.eigent.ai/x
8
30
166
12,987
🐫 Using CAMEL's Role-Playing Scraper for a Report & Knowledge Graph on the Turkish Shooter from the 2024 Paris Olympics! 🥇
Dive into web scraping, multi-agent role-playing, and knowledge graph construction in one comprehensive example. 🕵️♂️🤖
⭐ Highlights:
- 🛠️ Setup: Install @CamelAIOrg & configure API keys.
- 🔍 Web Search: Search and get web links with @DuckDuckGo.
- 🔥 Web Scraping: Get LLM-friendly markdown content from the URLs with @firecrawl.
- 🧩 Chunking: Organize data with @UnstructuredIO.
- ✨ Embedding: Embed data with @MistralAI Embedding.
- 🗂️ Vector database: Retrieve embeddings with @Qdrant_engine.
- 🐫 Info Retrieval: Aggregate data using @CamelAIOrg's RAG.
- 🧠 Knowledge Graph: Generate with @CamelAIOrg's KG agents and store with @neo4j.
- 🤖 Multi-Agent Role-Playing: Automate tasks with AI agents using @MistralAI models.
- 👀 Agent Monitoring: Use @AgentOpsAI to manage and oversee operations.
👉 Try it for yourself here: colab.research.google.com/dr…
🐫 Build your own multi-agent system here: github.com/camel-ai
3
35
158
28,346
🚀 GraphRAG: Boost Retrieval Accuracy and RAG Performance!
Check out the diagram and tech stack below to see how this works.
🔹@CamelAIOrg: A multi-agent system framework that facilitates the workflow and provides agents such as Knowledge Graph Agent and Chat Agent.
🔹@MistralAI: Provides the LLM Mistral Large 2 and an embedding model, Mistral Embed.
🔹@neo4j: Knowledge graph database.
🔹@qdrant_engine: Vector database.
🔹@UnstructuredIO: Gets the data ready for RAG.
Thanks to @ttokzzzzz and @guohao_li for their contributions to this. 🤝
Thanks to @sophiamyang and @MistralAI for collaborating on this project. 🤝
Try it out for yourself via this cookbook: github.com/mistralai/cookboo…
GraphRAG with @MistralAI, @CamelAIOrg, and @neo4j:
- Use Mistral Large 2 to extract and structure knowledge graph from a given content source, and store this information in a Neo4j graph database.
- A hybrid approach: combining vector retrieval and knowledge graph retrieval, to query and explore the stored knowledge.
- This hybrid Graph + vector RAG approach increases retrieval accuracy and RAG performance.
Thank you @ttokzzzzz @guohao_li for contributing this great example to Mistral cookbook. Link in 🧵:
2
45
156
55,904
🦉 Owl by CAMEL AI just took flight！
🌟Less than 24 hours — 2.3K stars and counting
🏆#1 on GAIA, outperforming Hugging Face’s Open Deep Research
A fully open-source general AI agent
🚀Next up:
OWL developer meeting & Discord community launching soon！
Join the action and check it out today! github.com/camel-ai/owl
1
30
143
18,474
OWL now speaks MCP.
A universal interface to let OWL agents access external tools.
(securely, flexibly, and with zero rewrite)
One config.
Multiple tools.
Fully autonomous task execution.
Read the detailed blog to get started
camel-ai.org/blogs/owl-mcp-t…
2
28
142
5,057
Try GraphRAG with CAMEL.
🚀 GraphRAG: Boost Retrieval Accuracy and RAG Performance!
Check out the diagram and tech stack below to see how this works.
🔹@CamelAIOrg: A multi-agent system framework that facilitates the workflow and provides agents such as Knowledge Graph Agent and Chat Agent.
🔹@MistralAI: Provides the LLM Mistral Large 2 and an embedding model, Mistral Embed.
🔹@neo4j: Knowledge graph database.
🔹@qdrant_engine: Vector database.
🔹@UnstructuredIO: Gets the data ready for RAG.
Thanks to @ttokzzzzz and @guohao_li for their contributions to this. 🤝
Thanks to @sophiamyang and @MistralAI for collaborating on this project. 🤝
Try it out for yourself via this cookbook: github.com/mistralai/cookboo…
1
18
127
28,183
Your agents. Your tools. One protocol.
𝐎𝐖𝐋 𝐣𝐮𝐬𝐭 𝐠𝐨𝐭 𝐦𝐨𝐫𝐞 𝐞𝐱𝐭𝐞𝐧𝐬𝐢𝐛𝐥𝐞 𝐰𝐢𝐭𝐡 𝐌𝐂𝐏.
No more embedded tool logic
Just define your MCPServers and go
Watch how OWL now talks to real-world services through MCP 👇
3
12
127
5,964
𝐘𝐨𝐮𝐫 𝐂𝐀𝐌𝐄𝐋-𝐀𝐈 𝐀𝐠𝐞𝐧𝐭 𝐢𝐬 𝐧𝐨𝐰 𝐚 𝐭𝐨𝐨𝐥.
You can now serve any CAMEL-AI agent as an MCP server, letting external clients like Claude Desktop call it directly.
Here’s how it works:
🧵 Day 2/7
1
18
113
12,295
🦉𝐎𝐖𝐋 × @WhatsApp 𝐌𝐂𝐏 𝐢𝐬 𝐬𝐥𝐢𝐜𝐤.
You don’t need to define a workflow.
You don’t even mention which tool to use.
Just a plain message.
→ OWL connects to WhatsApp via the MCP toolkit
→ Roleplay agents infer the intent
→ Picks the right tool and replies — all in real time
𝐌𝐂𝐏 𝐦𝐚𝐤𝐞𝐬 𝐢𝐭 𝐦𝐨𝐝𝐮𝐥𝐚𝐫. 𝐎𝐖𝐋 𝐦𝐚𝐤𝐞𝐬 𝐢𝐭 𝐢𝐧𝐭𝐞𝐥𝐥𝐢𝐠𝐞𝐧𝐭.
3
14
112
11,287
Introducing Project Loong 🐉
Blog: camel-ai.org/blogs/project-l……
• Our structured approach to generating and validating synthetic data for enhanced model performance.
• Modular design that integrates synthetic data generation with semantic verification
• Multi-agent framework ensuring accuracy and consistency
• Empowering domain-specific models with reliable reasoning signals
4
29
111
5,391
🦉 OWL has surged to 8.8K+ stars in just 4 days and growing！
- Mar 7: Open-sourced OWL
- Mar 9: Launched a web UI for easier interaction
- Mar 11: Added MCPToolkit, FileWriteToolkit & TerminalToolkit for enhanced agent capabilities
- Mar 12: Will be kicking off our Community Call for Use Cases
Explore, build, and push the limits of open-source AI with OWL. 🚀
3
7
103
8,065
𝐓𝐰𝐨 𝐲𝐞𝐚𝐫𝐬 𝐨𝐟 𝐂𝐀𝐌𝐄𝐋. 🎂🐫
From a simple framework to a growing agent ecosystem.
All thanks to you.
We’re kicking off our birthday week with
5 straight days of new launches.
Day 1 starts here →
Scaling Environments for Agents
camel-ai.org/launchweek-envi…
Stay tuned.
This week’s going to be 𝐁𝐈𝐆.
10
14
96
8,595
Turn slides and documents into rich study notes in seconds with @Microsoft MarkItDown MCP and @CamelAIOrg agents
→ Auto-convert PPTs, PDFs, DOCXs into clean, structured Markdown
→ Harness Gemini OCR for pixel-perfect text extraction
→ Camel roleplay agents dissect, summarize, and format your notes
→ Fully modular, end-to-end workflow
100% open source
6
8
87
7,299
🔥 Building an AI-powered Discord bot with Agentic RAG has never been easier!
Learn how 🐫 @CamelAIOrg, 🔥 @firecrawl & 👨🚀 @qdrant_engine work together to create the ultimate customer service @discord bot powered by @Alibaba_Qwen or @MistralAI.
From web scraping to qdrant's vector database - we've got your community covered with this bot!
👇 Full architecture breakdown in the figure below.
🔗 Check out the colab: colab.research.google.com/dr…
8
27
84
13,112
Exciting New Feature: MCP Toolkit in CAMEL-AI! 🐫
We're thrilled to announce the latest addition to CAMEL-AI: the Model Context Protocol (MCP) Toolkit for seamless integration with external tools!
Key Capabilities:
✅ Connect to external tools via MCP standard
✅ Support for both stdio and SSE connection modes
✅ Dynamic function generation from tool definitions
✅ Handle various content types (text, image, embedded resources)
✅ Seamless integration with CAMEL-AI's agent architecture
This toolkit creates a powerful bridge between your AI agents and external tools and services - dramatically expanding the capabilities of your CAMEL-AI applications!
1
8
76
11,519
A Google Forms MCP Server powered by @CamelAIOrg Agents.
Generate and manage forms using natural language.
Fully modular, tool-call driven, and open source.
Shoutout to @AdarshPandey355 for building something seriously cool 🔧
Adarsh Pandey
4
11
72
3,187
CAMEL-AI now supports @Kimi_Moonshot Kimi K2
What’s special about Kimi K2?
✔️ Trillion-parameter Mixture-of-Experts (MoE) architecture (32B active per inference)
✔️ Purpose-built for autonomous tool use, code execution, and workflow orchestration
✔️ Superior benchmark performance:
65.8% on SWE-bench Verified (agentic coding)
53.7% on LiveCodeBench
✔️ 128K token context for complex, multi-step tasks
✔️ 5x cheaper than similar proprietary models
✔️ Open-source and subscription-free
With Kimi K2, CAMEL-AI users gain next-level multi-agent capabilities, advanced code generation, and seamless handling of long-context, autonomous tasks.
Try Kimi K2 with CAMEL-AI today!
3
11
74
11,125
We are so grateful to our CAMEL-AI community contributors for laying the foundation over the past two years. Today, Eigent.AI is finally coming out of stealth and meeting the world!
Introducing Eigent — the first multi-agent workforce on your desktop.
Eigent is a team of AI agents collaborating to complete complex tasks in parallel. It is your long-term working partner with fullly customizable workers and MCPs.
Public beta available to download for MacOS, Windows. 100% open-source on Github. Comment for 500 extra credits.
7
12
75
17,426
Introducing Multimodal Support with Ollama in CAMEL-AI!
We're excited to announce native support for Ollama multimodal models in CAMEL-AI! Now you can harness the power of vision-language models like llava-phi3 right from your CAMEL agents.
Key Features:
- Seamless image processing
- Native Ollama integration
- Simple API for multimodal inputs
- Easy-to-follow example implementation
This opens up exciting possibilities for vision-enabled AI agents and multimodal conversations!
1
9
63
5,578
1/ For no cost, this notebook offers the following capabilities:
- Generates high-quality CoT data.
- Fine-tunes @Alibaba_Qwen's Qwen2.5-1.5B using @UnslothAI.
- Uploads the model & data to @huggingface.
1
8
62
4,176
⚡ Build the fastest answering customer service AI-powered Discord bot!
Learn how 🐫 @CamelAIOrg, 🔥 @firecrawl & 👨🚀 @qdrant_engine work together to create the ultimate customer service @discord bot powered by @Alibaba_Qwen's Qwen 2.5 70b and Qwen 2.5 Coder 32b models severed by ⚡ @SambaNovaAI.
@SambaNovaAI is the clear leader in output speed for using the Qwen 2.5 models; check the quoted post for a full breakdown.
👇 Full architecture breakdown in the figure below.
🔗 Run the code via our cookbook: docs.camel-ai.org/cookbooks/…
Benchmarks of providers of Qwen2.5, a leading open-source model family 📊
@alibaba_cloud's Qwen2.5 family of models includes Qwen2.5 72B, Qwen2.5 Coder 32B and a range of smaller models including 1.5B and 0.5B models for ‘edge’ use-cases.
Qwen2.5 72B, the flagship model, is competitive in intelligence evaluations with frontier models including Llama 3.3 70B, GPT-4o and Mistral Large 2.
Despite its smaller size, Qwen 2.5 Coder 32B achieves comparable performance in coding benchmarks like HumanEval to frontier models. Its size and capabilities position it well to support developers with fast code generation and emerging use-cases such as coding agents that require multi-step inference to autonomously develop features and applications.
Amongst providers, @SambaNovaAI is the clear leader in output speed, delivering ~225 output tokens/s on Qwen2.5 72B, and 566 output tokens/s on Qwen 2.5 Coder 32B in our coding workload benchmark.
@nebiusai , @DeepInfra , @hyperbolic_labs and @togethercompute are also offering the model(s) and all at prices significantly cheaper than comparable proprietary models such as GPT-4o.
Links to our live benchmarks of Qwen2.5 on Artificial Analysis below 👇
3
11
61
9,289
OpenAI’s Agent SDK release aims for a unified framework, but it’s clearly impossible. On the other hand, Anthropic’s MCP makes it so that whether the frameworks are unified or not doesn’t really matter — it is clear that OpenAI is losing the battle 🫡
7
7
62
5,265
📢 Mastering Multi-Agent Systems:🐫 CAMEL-AI Workshop & Hackathon
Join us for an exciting event on October 25th at @UniofOxford, featuring a workshop and a hackathon focused on multi-agent systems!
We will have talks from speakers such as
→ @philiptorr (Professor @UniofOxford & Chief Scientific Advisor @Eigent_AI)
→ @guohao_li (Founder @Eigent_AI & @CamelAIOrg)
You’ll learn how to set up your first multi-agent system—no coding experience needed. 💻
The workshop covers agent basics, prompts, tools, and agentic structures. In the hands-on Workforce Workshop, you'll create multi-agent systems for real-world use cases. 🤖
Then, put your new skills to the test in our hackathon! Compete with teammates to build a multi-agent system project to win prizes. 🏆
(Don't worry if you don't have a team in mind. We will host some networking activities in the morning!)
A big thank you to our partners & sponsors: @OxfordAI, Oxford Founders Society, @MistralAI, @firecrawl, @qdrant_engine, @SambaNovaAI, @AgentOpsAI, Chunkr (@lumina_ai_inc) & @Eigent_AI.
Register here👉 lu.ma/nktfdtqj
We look forward to seeing what you build!
10
10
58
25,681
📢 We've just added support for the Qwen models from Tongyi @Alibaba_Qwen (Meet Qwen builders: @JustinLin610, @huybery ) in the 🐫 CAMEL framework!
💥 By integrating the Qwen series models, including Qwen2.5-Coder (specialized in code generation and repair), Qwen-max (a high-performance model), Qwen-plus (optimized for general-purpose performance), Qwen-turbo (focused on execution speed), and Qwen-long (designed for long-text processing), the diversity of the CAMEL framework has been significantly enhanced, enabling the platform to meet the diverse needs of users. 😆
Thanks to our contributor MuggleJinx for this significant contribution!
🤝 Explore more here: github.com/camel-ai/camel/pu…
3
9
56
10,409
𝐎𝐖𝐋 × 𝐌𝐂𝐏: 𝐂𝐨𝐦𝐩𝐥𝐞𝐱 𝐓𝐚𝐬𝐤𝐬, 𝐎𝐧𝐞 𝐑𝐮𝐧 ⚡
Let OWL’s role-playing agents choreograph calls to multiple MCP servers, dynamically pick the best tool for each sub-task, and execute a full multi-step workflow in one seamless run.
Time to build 👇
🧵 Day 4/7
1
15
55
2,846
Smart Reasoning in CAMEL-AI with Rejection Sampling!
We're excited to showcase a powerful enhancement in 🐫 CAMEL-AI: Rejection Sampling for High-Quality Reasoning!
Key Features:
✅ Generates multiple reasoning traces simultaneously
✅ Smart evaluation of each candidate solution
✅ Automatic selection of highest-quality reasoning
✅ Quality threshold filtering for reliable outputs
This breakthrough ensures your AI agents produce more reliable and higher-quality reasoning, elevating the standard of automated problem-solving!
3
7
53
11,066
From PDF to polished poster, powered by agents.
We spotted this awesome Paper2Poster project that turns academic papers into visual posters using CAMEL & OWL.
One of the most useful AI-for-research applications we’ve seen lately.
Brilliant work by the team.
2
6
53
3,225
🎙️ Improving Reward Models with Synthetic Critiques - Zihuiwen Ye 📢
Thanks to @Daniella_yz for this amazing talk about the paper “Improving Reward Models with Synthetic Critiques” by @UniofOxford and @cohere.
Thanks to everyone who joined and participated in the Q&A! We will host more events like this soon in the 🐫 CAMEL-AI Discord.
- 📰 Check out the paper: arxiv.org/abs/2405.20850
- 🔗 Join the Discord: discord.gg/8zQTBNqf97
9
51
23,593
The best open source general AI agent is on 🔥!
Excited to share our new project 🦉OWL - an open-source alternative to Manus AI with 6K+ stars and climbing…
🦉OWL: github.com/camel-ai/owl
Everyone’s talking about Manus AI (an agent that can research, browse, code, and automate tasks.)
But there’s a catch:
→ All the above agents cost $20 to $200 a month
→ Manus AI is available via an invite code only
OWL by CAMEL-AI is an opensource project that’s breaking records:
⭐ 6K+ GitHub stars in 2 days
🏆 Ranks #1 on the GAIA Benchmark among open-source projects with a 58.18 avg score!
📌 Specialized Multi-Agent Collaboration
↳ AI user agents break down complex tasks
↳ Assistants agents create detailed execution strategies and interact with tools and tool agents
↳ Tools / Tool agents interface with external services and APIs
📌 Real-World Capabilities
↳ Autonomous research, coding, and web browsing
↳ Runs locally on your machine for privacy
↳ Supports top LLMs (easily integrated with GPT-4O, Claude 3.7, Gemini, Mistral, DeepSeek , Qwen, Ollama, Groq more)
↳ Automates complex workflows, from research to execution
I
📌 Simple Setup
↳ Fast installs via conda or uv
↳ Docker support for easy deployment
↳ Straightforward config with TOML files
No restrictions.
No invites.
No paywalls.
Just powerful multi-agent collaboration.
AI autonomy shouldn’t be locked behind paywalls.
4
7
47
8,354
Weekly 🦉 OWL Update – We’re 1 week old!
Big week for OWL! Our tech team has been grinding and here’s what’s new:
🗓 Mar 15 (Sat)
Major UI revamp, smoother and more stable! 🔧
OWL Agent optimization = better efficiency & performance!
SearchToolkit now integrates Baidu for easier searches!
🗓 Mar 12 (Wed)
Bocha search added to SearchToolkit + support for Volcano Engine models!
Azure & OpenAI-compatible models now have improved structured output & tool use!
🗓 Mar 11 (Tue)
New toolkits: MCP, FileWrite & Terminal—more power for OWL Agent!
MCP protocol standardizes interactions between AI models, data, and tools!
🗓 Mar 9 (Sun)
Web-based UI is live—better interactions!
🗓 Mar 7 (Fri)
🦉 OWL is now open source – check out the code!
🗓 Mar 3 (Mon)
OWL scored 58.18 on the GAIA benchmark, leading open-source frameworks!
We’re also looking for more 🦉 OWL use cases! Got one? Shoot us a DM!
4
7
45
2,815
𝐎𝐖𝐋: 𝐓𝐡𝐞 #𝟏 𝐎𝐩𝐞𝐧-𝐒𝐨𝐮𝐫𝐜𝐞 𝐌𝐮𝐥𝐭𝐢-𝐀𝐠𝐞𝐧𝐭 𝐂𝐨𝐥𝐥𝐚𝐛𝐨𝐫𝐚𝐭𝐢𝐨𝐧 𝐅𝐫𝐚𝐦𝐞𝐰𝐨𝐫𝐤
𝐑𝐚𝐧𝐤𝐞𝐝 #𝟏 on GAIA Benchmark! 🏆
𝟐.𝟑𝐊+ 𝐆𝐢𝐭𝐇𝐮𝐛 𝐬𝐭𝐚𝐫𝐬 in just 1 day! ⭐
Here’s why everyone’s talking about OWL 👇
9
13
44
21,228
🎙️"Agent Workflow Memory" - Zora Zhiruo Wang 📢
Thanks to @ZhiruoW for this amazing talk about the paper "Agent Workflow Memory" by @CarnegieMellon & @MIT.
Thanks to everyone who joined and participated in the Q&A! We will host more events like this soon in the 🐫 CAMEL-AI Discord!
- 👉 Watch the whole talk: piped.video/watch?v=sr9Np5xQ…
- 📰 Check out the paper: arxiv.org/abs/2409.07429
- 🐫 Join the Discord: discord.gg/8zQTBNqf97
2
8
45
6,797
CAMEL-AI Ambassador Program is opening up again!
Come join us:
> Collaborate on hands-on AI projects
> Connect & work with our core team
> Score early access to upcoming CAMEL-AI products
> Even get a chance to join our team down the road
Apply now: forms.gle/JXnCh72DSuAHQ7Hh8
7
4
43
1,964
🚀 Effortlessly Supercharge Your Model with Agentic SFT Data Generation & Fine-Tuning
In this notebook, you will learn 🐫 @CamelAIOrg, 🦥 @UnslothAI & 🔥 @firecrawl make an excellent team.
Learn how to generate data and fine-tune open-source models of your choice to significantly improve their ability to handle content directly from web pages. 🌐
👇Try it out for free with these models:
- @Alibaba_Qwen's Qwen2.5-7B: docs.camel-ai.org/cookbooks/…
- @MistralAI's mistral-7b-instruct-v0.2-bnb-4bit: docs.camel-ai.org/cookbooks/…
- @Meta's tinyllama-bnb-4bit: docs.camel-ai.org/cookbooks/…
2
14
44
6,869
When AutoGen released another boring idea...
🐫 CAMEL-AI.org just announced 🏝️ OASIS with one million agents for social simulations.
3
10
42
3,512
CAMEL-AI Now Integrates with @NetMindAI Platform!
Key Features:
- Unlock quicker and more efficient AI processing with NetMind’s robust infrastructure.
- Diverse Model Options: Compatible with leading models like DeepSeek, Llama 4, and Qwen 3.
- Easily scale your AI applications with Netmind’s flexible platform.
This integration expands CAMEL-AI’s capabilities, providing users with more options to build and deploy powerful AI solutions.
2
8
43
2,951
Turn your PDFs into clean, structured summaries with a single-agent pipeline.
We just integrated @MistralAI OCR into CAMEL-AI, so you can extract and summarize complex, multilingual documents out of the box.
All wrapped in one ready-to-use cookbook.
📘 docs.camel-ai.org/cookbooks/…
6
41
5,961
DeepSeek R1 is amazing for data generation!
Want structured, high-quality mathematical reasoning datasets?
With 🐫 CAMEL-AI + 🐋 @DeepSeek R1, distilling detailed thought processes is effortless!
Follow this step-by-step guide to get started!
[ Bookmark to use later 🧵]
1
8
39
3,291
🐫 CAMEL’s Role-Playing supercharges AI agents by enabling them to tackle complex tasks autonomously. 🚀
Role-Playing consists of the following key parts:
• 🔍 Task Prompt: A task can be as simple as a single idea, kickstarted by an inception prompt that sets the objective.
• 💡 AI User: Acts as the strategist, responsible for giving detailed instructions that guide the overall task.
• 🤖 AI Assistant: Serves as the problem-solver, providing precise solutions to the AI User’s instructions.
🔧 How it works: Roles are assigned to the agents. Then, both agents interact in a structured dialogue where the AI User plans and the AI Assistant executes with any tool you provide.
In the diagram from the tweet below, you’ll see where agents powered by @MistralAI models and equipped with numerous tools use Role-Playing for creating reports & knowledge graph generation.
🔗 Try it out for yourself via the cookbook in the tweet below.
Mutli-agent role-playing and knowledge graph construction with @MistralAI, @CamelAIOrg, @DuckDuckGo, @firecrawl, @UnstructuredIO, @qdrant_engine, @neo4j, @AgentOpsAI!
- 🔍Web Search: Search and get web links with @DuckDuckGo.
- 🔥 Web Scraping: Get LLM-friendly markdown content from the URLs with @firecrawl.
- 🧩 Chunking: Organize data with @UnstructuredIO.
- ✨ Embedding: Embed data with @MistralAI Embed.
- 🗂️ Vector database: Retrieve embeddings with
@Qdrant_engine.
- 🐫 Info Retrieval: Aggregate data using
@CamelAIOrg's RAG.
- 🧠 Knowledge Graph: Generate with
@CamelAIOrg's KG agents and store with @neo4j.
- 🤖 Multi-Agent Role-Playing: Automate tasks with AI agents using @MistralAI Mistral Large 2.
- 👀 Agent Monitoring: Use
@AgentOpsAI to manage and oversee operations.
Big shout out to @guohao_li @ttokzzzzz for contributing this amazing example to Mistral cookbook! Link in 🧵
9
39
3,220
Connecting your agents to the real world just got easier.
Explore how to make your CAMEL agent as MCP client, and unlock powerful real-world use cases.
It’s like giving your agents USB-C access to the world.
🧵 Day 1/7
1
13
40
5,419
🔥 Multi-agent is the future, and CAMEL AI is leading the way!
If you're as excited about this as we are, smash that ⭐ on GitHub and join the revolution.
🔗 GitHub Repo github.com/camel-ai/camel
4
35
5,908
Enhancing Memory Management: Timestamp Support in CAMEL-AI!
We're excited to introduce timestamp tracking in VectorDBMemory, ensuring precise chronological data retrieval in agent conversations!
Key Improvements:
✅ Accurate conversation order preservation
✅ Enhanced context creation with temporal awareness
✅ Improved memory block management
✅ New clear() method for memory reset
This update prevents data retrieval confusion and makes agent interactions more reliable!
1
2
37
2,048
From just one prompt, 🦀 CRAB set a timer on my phone based on a message it found on Slack. 🤖
The prompt used: "Tianqi Xu sent me a reminder on Slack on my PC. Please set and start a timer on my phone accordingly."
See the sped-up demo below to better understand how 🦀 CRAB works.
Try it out for yourself: github.com/camel-ai/crab
Introducing 🦀 CRAB: Cross-environment Agent Benchmark for Multimodal Language Model Agents
🦀 CRAB provides an end-to-end and easy-to-use framework to build multimodal agents, operate environments, and create benchmarks to evaluate them, featuring three key components:
- 🔀 Cross-environment support - agents can operate tasks in 📱 Android and 💻 Ubuntu.
- 🕸️ Graph evaluator - provides a fine-grain evaluation metric for agents.
- 🤖 Task generation - composes subtasks to automatically generate tasks.
By connecting all devices to agents, 🦀CRAB unlocks greater capabilities for human-like tasks than ever before.
Use 🦀 CRAB to benchmark your multimodal agents!
- 👨💻 Check out the repository: github.com/camel-ai/crab
- 📝 Read the paper: arxiv.org/abs/2407.01511
- 🌐 Find out more via the project page: crab.camel-ai.org/
- 🐫 Join our community: discord.gg/8zQTBNqf97
4
34
2,839
ASI has arrived internally, 🦀 CRAB agent exits VIM successfully.
It is so over; we are seeing a VLM agent completing a task that required 14 steps.
Introducing 🦀 CRAB: Cross-environment Agent Benchmark for Multimodal Language Model Agents
🦀 CRAB provides an end-to-end and easy-to-use framework to build multimodal agents, operate environments, and create benchmarks to evaluate them, featuring three key components:
- 🔀 Cross-environment support - agents can operate tasks in 📱 Android and 💻 Ubuntu.
- 🕸️ Graph evaluator - provides a fine-grain evaluation metric for agents.
- 🤖 Task generation - composes subtasks to automatically generate tasks.
By connecting all devices to agents, 🦀CRAB unlocks greater capabilities for human-like tasks than ever before.
Use 🦀 CRAB to benchmark your multimodal agents!
- 👨💻 Check out the repository: github.com/camel-ai/crab
- 📝 Read the paper: arxiv.org/abs/2407.01511
- 🌐 Find out more via the project page: crab.camel-ai.org/
- 🐫 Join our community: discord.gg/8zQTBNqf97
1
6
35
4,894
Ever wish your slides could write themselves?
We built 𝐩𝐩𝐭𝐱_𝐭𝐨𝐨𝐥𝐤𝐢𝐭, a new CAMEL-AI toolkit that lets you generate full PowerPoint decks using just a prompt.
Works with @OpenAI, @MistralAI, or any open-source model.
THREAD 🧵
[1/]
2
3
33
2,333
🤖 @Agora_Protocol is a cross-platform, dead-simple protocol for efficient communication between LLM agents. 💬
Check out the @Gradio demo; it shows how CAMEL agents can communicate cheaply with other agents: huggingface.co/spaces/agora-…
We've just released a @Gradio demo for Agora🏛️! Try it yourself to see how Agora makes communication between agents way cheaper.
huggingface.co/spaces/agora-…
2
8
32
14,223
Introducing PythonVerifier: Robust Code Verification in CAMEL-AI!
We're excited to announce PythonVerifier - a powerful new addition to our verification framework that enables automated Python code testing and validation!📷
Key Features:
- Secure code execution in isolated virtual environments
- Allows for package installation in the virtual env to run any code.
- Intelligent expression vs code block handling
- AST-based evaluation and comparison
- Comprehensive error handling
- Execution timeout protection
This tool makes Python code verification seamless and secure in CAMEL-AI workflows!
2
5
33
4,273
Scaling Environments for Agents Launch Week
Day 4: Project OASIS 🏝️
The world’s first multi-agent social media simulation platform capable of simulating one million agents.
Blog 👉 Automation or Simulation? - The Biggest Potential of Multi-agent systems: camel-ai.org/blogs/project-o…
3
7
34
1,083
We're launching an open-access 🐫 AI Agent Course Project, designed to narrow the gap between cutting-edge AI research and classroom instruction.
Researchers, educators, and builders: come build it with the community, not just for it.
🔗 Learn more: camel-ai.github.io/agents-co…
📝 Apply to contribute: eigent-ai.notion.site/206511…
1
7
35
1,721
🦀 CRAB aims to become a general-purpose agent benchmark framework for Multimodal Language Model (MLM) agents. 🤖
We present to you CRAB Benchmark-v0, developed using the CRAB framework, which includes 100 tasks across 2 environments (Ubuntu and Android) and was tested with 4 different MLMs (GPT-4o, GPT-4 Turbo, Gemini 1.5 Pro and Claude 3 Opus) under 3 distinct communication settings. 🚀
This is just the start, 🦀 CRAB will be a long-term open-source effort. More exciting new features, benchmarking results and research projects to come.
Watch this video for more details about the project. 👇
Introducing 🦀 CRAB: Cross-environment Agent Benchmark for Multimodal Language Model Agents
🦀 CRAB provides an end-to-end and easy-to-use framework to build multimodal agents, operate environments, and create benchmarks to evaluate them, featuring three key components:
- 🔀 Cross-environment support - agents can operate tasks in 📱 Android and 💻 Ubuntu.
- 🕸️ Graph evaluator - provides a fine-grain evaluation metric for agents.
- 🤖 Task generation - composes subtasks to automatically generate tasks.
By connecting all devices to agents, 🦀CRAB unlocks greater capabilities for human-like tasks than ever before.
Use 🦀 CRAB to benchmark your multimodal agents!
- 👨💻 Check out the repository: github.com/camel-ai/crab
- 📝 Read the paper: arxiv.org/abs/2407.01511
- 🌐 Find out more via the project page: crab.camel-ai.org/
- 🐫 Join our community: discord.gg/8zQTBNqf97
13
32
4,782
🚀 CAMEL-AI MCP Launch Week starts Monday!
We’ll be sharing some of the most exciting updates and features we’ve been working on. Stay tuned! 👀🔥
1
6
33
1,383
Ever wanted to create an intelligent AI agent? Our latest step-by-step tutorial teaches you How to Construct a 🐫CAMEL Agent to Build a Laptop Recommendation System using @deepseek_ai R1 in @aimlapi setup to full interaction! 💡
🔧 What you’ll learn:
✅ Configure DeepSeek R1 in CAMEL for reasoning & insights
✅ Build an interactive ChatAgent with memory & tools
✅ Process datasets & provide structured AI-driven recommendations
✅ Learn through a real-world case study: a Laptop Recommendation System that analyzes specs & suggests the best choices!
👨💻 Perfect for AI developers & researchers! Ready to build your own AI-powered assistant? Start here:
📖 Tutorial: Colab Notebook colab.research.google.com/dr…
#AI #CAMELAI #DeepSeek #AIML #MachineLearning #AIrecommendation #MultiAgentSystems
1
9
33
5,099
Thanks to the @googleaidevs team for featuring 𝐎𝐖𝐋🦉
Exciting to see OWL in action with Gemini 2.5 Pro for automating complex tasks.
🦉 Automate complex tasks using Gemini 2.5 Pro and @CamelAIOrg’s OWL (Optimized Workforce Learning), an open-source multi-agent collaboration framework that works together like a real-world project team.
ALT System architecture diagram for OWL showing a User Query interacting with an AI User Agent and Assistant Agent which communicate with each other. These agents access specific Actor Agents (Web, Search, Coding, Document) which utilize tools from a Tools Pool (Browser, Multimodal Tools, Document Processing Tool, Code Executor). The diagram is titled "OWL | System Architecture" and includes the subtitle "Optimized Workforce Learning for General Multi-Agent Assistance in Real-World Task Automation", a repository link, and the CAMEL-AI logo.
1
4
29
1,856
Exciting Update: CAMEL-AI Now Integrates with @mem0ai for Cloud Memory!
We're thrilled to announce native Mem0 integration in CAMEL-AI, bringing powerful cloud-based memory capabilities to your AI agents!
Key Features:
- Cloud-persistent chat history
- Seamless session synchronization
- Simple setup & configuration
- Reliable memory storage
- Production-ready scalability
This integration enables your agents to maintain consistent memory across sessions and deployments.
1
3
31
3,915
Seamless Notion automation powered by MCP × OWL
Watch how our agents:
• Read your Notion workspace in real time
• Parse and understand page content via MCP
• Execute updates, queries, and new entries
Just a plain prompt, and your Notion flows on autopilot.
100% open source
2
5
30
3,963
CAMEL-AI's Workforce can now be exported as an MCP Server
MCP Clients like Claude, @cursor_ai and others can now call into a CAMEL-powered agent workflow
All with just a few lines of code.
Scalable, modular, and interoperable by design.
🧵 Thread below ↓
3
4
30
1,958
AI Agents are powerful, but it still needs us. 🤝
Let’s talk about 𝗛𝘂𝗺𝗮𝗻-𝗶𝗻-𝘁𝗵𝗲-𝗟𝗼𝗼𝗽 𝗔𝗜 (𝗛𝗜𝗧𝗟) ↓
3
2
30
6,223
Let your code speak!
CAMEL Agent as an MCP client + @gitingest MCP server transform any GitHub repo into a chat-ready AI that answers your questions and streamlines your workflow.
Try it now! 🚀
3
5
30
6,187
We just crossed 𝟏𝟎𝐊 stars on GitHub! ⭐️
Appreciate all the love from the community
🐫 github.com/camel-ai/camel
3
28
3,139
Introducing Enhanced Bing Search in 🐫 CAMEL-AI: Bridging Global Knowledge!
We’re excited to announce a powerful addition to CAMEL-AI’s search capabilities: Robust Bing Search Integration!
Key Features:
- Smart timeout handling (10s) for reliable performance
- Rich search results with titles, snippets, and URLs
- Optimized for both English & Chinese queries
- Clean error handling and response formatting
- Configurable result limits
This enhancement makes CAMEL-AI’s search toolkit even more versatile, ensuring your agents can access global information seamlessly and reliably!
1
2
26
860
CAMEL-AI Announces Partnership with Wolfram|Alpha – Powering a New Era of Computational Intelligence!
We’re thrilled to announce a new partnership with Wolfram|Alpha. Building on this collaboration, CAMEL-AI is launching Wolfram|Alpha Toolkit, now available within the CAMEL-AI framework. This marks an exciting step in our journey, hinting at a promising synergy between CAMEL-AI and Wolfram|Alpha’s powerful computational capabilities.
✨Key Features:
- Direct access to Wolfram|Alpha’s world-class computational intelligence
- Advanced capabilities for mathematical and scientific computation
- Seamless integration for complex query processing
This initial integration, focusing on the Wolfram|Alpha LLM API, empowers CAMEL-AI agents with robust analytical tools, enhancing their problem-solving capabilities.
We envision this as the beginning of deeper collaborations, with exciting future integrations planned to bring even more sophisticated Wolfram|Alpha functionalities to our agents.
3
7
27
2,228
ASI has arrived internally, 🦀 CRAB agent exits VIM successfully.
It is so over; we are seeing a VLM agent completing a task that required 14 steps.
2
5
26
2,204
CAMEL-AI just released the PyPI package for 🏝️ OASIS, enabling large-scale social simulations on social platforms with millions of agents.
Core steps to get started:
1. pip install camel-oasis
2. oasis.make() to create the social media environment
3. env.reset() to start the simulation
4. env.step(action) to let agents post, like, and interact
5. env.close() to end the simulation
Simulate digital societies like never before! 🚀🚀
2
6
28
2,142
𝐙𝐞𝐫𝐨-𝐬𝐡𝐨𝐭 𝐠𝐚𝐦𝐞𝐬 𝐚𝐫𝐞 𝐰𝐚𝐲 𝐭𝐨𝐨 𝐟𝐮𝐧 𝐭𝐨 𝐛𝐮𝐢𝐥𝐝.
OWL took a single prompt. Together AI’s LLaMA 4 turned it into a live Python Snake Game.
→ OWL assigns the task
→ Together AI runs inference
→ Code is written and executed live
@CamelAIOrg meets @togethercompute
Seamless, Open, Autonomous
4
4
28
3,159
[Scaling Environment for Agents]
Launch week day 2
Project Loong 🐉 is releasing！
camel-ai.org/launchweek-envi…
6
28
1,355
We’re launching CAMEL-AI MCP Hub: the definitive directory for discovering and comparing Model Context Protocol servers to power your AI agents.
Available now at mcp.camel-ai.org/
1
2
27
1,496
We showed the demo, now here’s the full guide 🛠️📲
Dropped a detailed blog on how to connect your CAMEL-AI agent to @WhatsApp using OWL + MCP.
Read it here ↓
camel-ai.org/blogs/mcp-serve…
4
26
786
Exciting Update in CAMEL-AI!
Introducing our new internal deduplication module contributed by Keli Wen - now with smart similarity detection using embeddings! 📷
📷 Key Features:
- Flexible embedding support
- Configurable similarity thresholds
- Efficient duplicate mapping
- Cosine similarity-based matching
Perfect for cleaning your training data and ensuring unique, high-quality content! 📷
👉 Check it out here: github.com/camel-ai/camel/pu…
3
25
13,966
🐫 CAMEL-AI with @ollama - Run AI Agents with Local Models Easily 🚀
This video shows you how to create a real estate chatbot powered by a CAMEL agent that runs @Meta's Llama 3 model locally via Ollama. 🤖
Thanks to @fahdmirza for making this great tutorial! 🤝
👉 Watch it here: piped.video/watch?v=IyieOity…
2
5
26
1,507
🚨 CAMEL-AI Live Talk Alert!
Don’t miss this talk by Wei Pang(Waterloo CS Master’s, incoming CUHK-SZ PhD) on Paper2Poster – the first public benchmark for automated academic poster generation!
Learn how PosterAgent turns 20+ page papers into sleek, editable posters for just $0.005, beating GPT-4o in multiple metrics.
⏰ 4th July 2025 17:00 BST / 9:00 PDT / 12:00 EDT
📍 Google meet
4
2
27
3,312
📢 We've just added the Workforce module in the 🐫 CAMEL framework!
Workforce is a system where multiple agents work together to solve tasks. 🤖🤖🤖
Workforce follows a hierarchical architecture. A workforce can consist of multiple worker nodes, and each of the worker nodes will contain one agent or multiple agents as the worker.
The worker nodes are managed by a coordinator agent inside the workforce, and the coordinator agent will assign tasks to the worker nodes according to the description of the worker nodes, along with their tool sets. ⚒️
Alongside the coordinator agent, there is also a task planner agent inside the workforce. The task planner agent will take the responsibility of decomposing and composing tasks, so that the workforce can solve the task step by step. 🤔
In the example bellow, you can use how a workforce works together to with agents that have different tools to plan a trip to Paris.
See the example 👉colab.research.google.com/dr…
Thanks to our contributors @Whale__Eye & yiyiyi0817 for this significant update. 🤝 Explore more here: github.com/camel-ai/camel/pu…
Find out more about Workforce in our docs: docs.camel-ai.org/key_module…
9
26
8,834
Scaling Environments for Agents Launch Week
Day 3: OWL + CRAB + MCP
We're bridging the last mile in agent automation.
Blog: camel-ai.org/blogs/the-new-e…
→ OWL: Multi-agent browser, code, and doc automation
→ CRAB: 120+ cross-device benchmark tasks
→ MCP: Plug-and-play external tools
2
7
25
1,289
📢 We've integrated @firecrawl into the 🐫 CAMEL framework!
Firecrawl empowers our agents with clean LLM-ready data that can be crawled from any website.
Thanks to Firecrawl 🔥 for collaborating on this integration.
Thanks to our contributor @ttokzzzzz for this contribution. 🤝 Explore more here: github.com/camel-ai/camel/pu…
We are excited to announce our new integration with @CamelAIOrg!
With Camel AI and Firecrawl you can quickly build multi-agent systems that use data from the web.
Excited to have this integration live and stay tuned for a tutorial coming soon 👀
8
26
5,204
CAMEL🐫 × @deepseek_ai 🐋 × @aimlapi (Day 3/7)
📄 Build a RAG system with CAMEL and DeepSeek
Upload a document & ask any question
Get accurate, document-based answers.
Reliable, evidence-backed answers in real-time.
Here’s how it works (THREAD🧵)
4
3
26
2,322
Chat with any document
Just upload, extract, and ask.
With @MistralAI OCR + @CamelAIOrg , you can turn scanned PDFs or images into fully interactive chat sessions.
Tables, text, even multilingual receipts? All parsed, structured, and ready to talk.
1
1
26
3,983
Build your own AI-powered Customer Service Discord Bot with Local Models! 🚀
In this cookbook you will learn how to:
- Harness @ollama to run @Alibaba_Qwen's QwQ 32B-preview locally.
- Set up a @discord bot for customer service bot using 🐫 @CamelAIOrg, 🔥 @firecrawl, and ✨🤖 @qdrant_engine.
👇 Full architecture breakdown in the figure below.
Try out the cookbook for yourself: docs.camel-ai.org/cookbooks/…
A massive thanks to our contributor @avery4869 working on this! 🤝
1
8
26
2,084
We’re hosting 𝗖𝗔𝗠𝗘𝗟-𝗔𝗜 𝗣𝗿𝗼𝗷𝗲𝗰𝘁 𝗠𝗲𝗲𝘁𝗶𝗻𝗴, and you’re invited!
We’ll be chatting about:
→ Updates on new integrations
→ A look at the next big features
→ Open floor for questions + feedback
→ How you can contribute and get involved!
🗓️ Today at 17:00 BST / 09:00 PT / 10:30 IST
📍 Link to join: discord.gg/nvTcrNUA?event=13…
Come hang out, bring your ideas, and let’s keep pushing open-source AI forward.
See you there! 🐫🦉
3
24
1,494
📢 We've integrated @openbb_finance into the 🐫 CAMEL framework!
The OpenBBToolkit class is for financial data access and analysis.💰
This includes stock quotes, market screening, economic indicators functionality, and ChatAgent integration for interactive analysis.
Special thanks to our contributor, anjieyang, for leading this implementation. 🤝
👉 Explore more here: github.com/camel-ai/camel/pu…
2
23
4,920
𝐋𝐚𝐮𝐧𝐜𝐡 𝐖𝐞𝐞𝐤 𝐬𝐭𝐚𝐫𝐭𝐬 𝐧𝐨𝐰.
Join us for the Launch Week Developer Meeting.
→ Recap of what we’ve built over the past 2 years
→ Our roadmap and vision going forward
→ Discussion on the Scaling environments for agents initiative
→ Open Q&A with the team + a look at EigentBot in action
🗓️March 31st (10:00 PDT / 18:00 BST / 22:30 IST)
🗺️ discord.gg/8zQTBNqf97
The future of agents is data-driven. Come shape it with us.
2
4
25
1,044
CAMEL-AI Now Supports OpenAI's GPT-4.1!
We're thrilled to announce that CAMEL-AI now fully supports OpenAI's latest GPT-4.1 models, including:
✅gpt-4.1
✅gpt-4.1-mini
✅ gpt-4.1-nano
This integration allows our autonomous agents to leverage these powerful new models with full million-token context windows!
1
2
25
1,033
Navigate to owl/webapp.py and get started..
Try it now 👉 github.com/camel-ai/owl
1
22
1,197
CAMEL-AI now supports @MistralAI Magistral Medium
✅ Powerful reasoning with traceable chain-of-thought
✅ Excels in multilingual logic (English, French, Arabic, more)
✅ Real-world performance: 73.6% on AIME2024
✅ 10x faster response speeds with Flash Answers
✅ Ideal for legal, finance, engineering, and creative workflows
Magistral Medium delivers high-fidelity, step-by-step reasoning for complex use cases, bringing a new level of transparency and accuracy to CAMEL-AI.
Try Magistral Medium now via CAMEL-AI!
2
4
24
2,433
Introducing Hybrid Retrieval in 🐫 CAMEL AI!
We’re excited to share our latest feature that makes RAG simpler and more effective!
🔍 What’s new?
• Seamless combination of semantic & keyword search
• Easy-to-use unified retrieval interface
• Better search results out of the box 🎯
No complex setup needed - just import and go! Perfect for both beginners and experts in RAG applications.
1
7
24
1,088
Moonshot AI Joins 🐫 CAMEL-AI!
CAMEL-AI has officially integrated @MoonshotAI, bringing state-of-the-art language models to our open-source framework!
This integration enables seamless access to Moonshot’s state-of-the-art models through CAMEL’s unified interface, making it easier to build sophisticated AI applications. Whether you’re developing chatbots, agents, or AI assistants, you can now leverage Moonshot’s powerful models with just a few lines of code!
Key features:
- Simple API integration
- Unified model interface
- Flexible configuration options
Try it out today!
👥Like, repost, and share if you found this helpful!
3
1
24
4,233