| Cobuilder | Raised | Earnings |
|---|---|---|
| Backer | Creators | Backed |
|---|---|---|
| Cobuilder | Backer | Spent | Bought |
|---|---|---|---|
$0.05 | 14.6k | ||
$0.1 | 41.3k |
| Backer | Creators | Backed |
|---|---|---|
| No backers yet | ||
| Cobuilder | Raised | Earnings |
|---|---|---|
| No creators yet | ||
Best experiment on the internet right now. DeepSeek and Qwen both up 100% on Hyperliquid 🚀
Alpha Arena update: - Qwen 3 Max and DeepSeek up almost 95% and 80%, respectively. - Claude now up +10% after being -20% a few days ago (impressive!) - Grok just broke even again - Gemini and GPT need a miracle LOL
Qwen going 10x long ETH is wild. Paying off handsomely. Deepseek crushing it
บอทเทรด จาก coinpilot ผมใช้มา 5 วัน Roi ดีมาก ผมเลือก Qwen กับ Deepseek ใครอยากลอง ใช้ Ref. ได้ลด fee 10% นะครับ เมื่อคืนตลาดเขียว qwen ปิดกำไรไป $8000 ต้องลองดูช่วงตลาดแดงว่าจะทำงานดีมั๊ย 👉 Copy top traders on Hyperliquid. iOS: https://apps.apple.com/hk/app/coinpilot/id6747697788 Android: https://play.google.com/store/apps/details?id=bot.coinpilot.app Save 10% off fees with code: 07177c
It’s day 8 for Alpha Arena. Six LLM’s got $10k in funding and are trading perps on hyperliquid. Qwen 3 Max is up 80%, Gemini 2.5 Pro & GPT-5 are both down close to 70%. They’re the two models with the most trades. This is a really interesting experiment because if TradFi assets all come onchain, that means autonomous agents will be able to access any crypto/tradifi assets. https://nof1.ai
we are on day 5 (i think?!) of Alpha Arena. Nof1 put together a benchmark designed to measure AI's investing abilities. it gave each model $10k of real money in real markets with identical prompts. right now neck and neck between Qwen 3 Max and DeepSeek. also... this is entirely run on Hyperliquid.
📍DeepSeek đánh bại Grok và OpenAI trong cuộc thi trading crypto 📌 Trong cuộc thi giao dịch crypto Alpha Arena vừa đc tổ chức, 6 mô hình AI được http://Nof1.ai sử dụng để so sánh hiệu suất: DeepSeek V3.1, Grok 4, Claude 4.5 Sonnet, GPT-5, Gemini 2.5 Pro và Qwen 3 Max — mỗi bên cầm $10K vốn thật để trade futures trên Hyperliquid, với update dữ liệu thị trường mỗi 2 phút. Đây là bài kiểm tra sớm xem AI có thể thật sự sống sót trong thị trường điều khiển bởi tâm lý, lòng tham và nỗi sợ hãi hay ko. 📌Kết quả sau 3 ngày: - DeepSeek biến $10K -> $14K — tương đương +35 ~ 40%. - Grok 4 bám ngay sau +35%. - GPT-5 và Gemini lỗ nặng (GPT-5 lỗ gần -40%). - Dữ liệu cho thấy DeepSeek long tới 83% tổng lệnh, giữ trung bình 21 giờ, leverage x12.5, đạt Sharpe 13.3 và drawdown ~15.5% -> DeepSeek chọn cách đánh có kỷ luật, kiên nhẫn hơn nhiều so với các đối thủ đảo vị thế liên tục. 📌Điều khác biệt giữa các model AI là ở hành vi giao dịch. Trong khi DeepSeek và Grok giao dịch ít, giữ lệnh lâu → ít chịu ảnh hưởng bởi biến động ngắn hạn, giảm rủi ro từ spread và phí giao dịch thì GPT-5 và Gemini vào ra như bot scalper, thua lỗ 1 phần đến từ slippage và funding. -> Có lẽ kết quả đã cho thấy mô hình phân tích tuyệt vời chưa chắc đã mang lại lợi nhuận. Thứ quyết định vẫn là risk management và cách giữ vị thế giữa một thị trường 24/7 liên tục biến động. Khi hàng loạt người dùng cùng một mô hình, cùng một phương pháp vào lệnh, lợi thế sẽ tự triệt tiêu. Do vậy dù bạn nghĩ rằng mình đã thông tường hết mọi bí kíp giao dịch thì số đông ngoài kia cũng nghĩ tương tự, vậy chiến thắng sẽ thuộc về ai?
AI models are being beaten to a pulp by the market 🙈 Update the situation (https://nof1.ai/) after 6 days of fighting, the number of AI models granted $10K profitable trade has halved AE: 1️⃣Qwen 3 Max: $13.5K (+35%) 2️⃣Deepseek chat v3.1: $12.3K (+23%) 3️⃣Claude Sonnet 4.5: $8.8K (-11%) 4️⃣Grok 4: $8.2K (-17%) 5️⃣Gemini 2.5 Pro: $3.8K (-62%) 6️⃣GPT 5: $2.8K (-72%) Among them, the model with the strongest fluctuations is called Grok 4. Grok 4's wallet once led with $15K but has now dropped straight to 4th place (-17%). Claude also fluctuated around the 3rd position (-11%). Only DeepSeek and Qwen 3 maintained their form, being the only 2 AIs with profits. Meanwhile, Gemini and Chat GPT were competing for the most negative position on the table as their accounts continued to probe the bottom from the beginning until now.
Chinese AI models — DeepSeek and Qwen — have reportedly outperformed ChatGPT and Grok in crypto trading performance. With AI systems now analyzing markets, executing trades, and optimizing strategies faster than humans, it raises a big question: 👉 Is AI officially changing the game in crypto trading?
📊 DeepSeek showed a 120% return in just nine days during an AI trading competition. Six neural networks started with equal funds and the same instructions. The goal is to maximize profit. The contest runs until November 3. Qwen 3 Max (Alibaba) is currently second, while OpenAI and Google DeepMind are down about 60%. An interesting experiment. But then again, would you really trust AI with your money?
Haven't seen anything as cool like this today DeepSeek just showed a 120% return in just nine days during an AI trading competition. Six neural networks started with equal funds and the same instructions. The goal is to maximize profit. The contest runs until November 3. Qwen 3 Max (Alibaba) is currently second, while OpenAI and Google DeepMind are down about 60%. An interesting experiment. But then again, AI is really taking over
DeepSeek showed a 120% return in just nine days during an AI trading competition. Six neural networks started with equal funds and the same instructions. The goal is to maximize profit. The contest runs until November 3. Qwen 3 Max (Alibaba) is currently second, while OpenAI and Google DeepMind are down about 60%. An interesting experiment. But then again, would you really trust AI with your money?
I have also written Deepseek, QWEN, and an AI trading competition system, and I am preparing to open source them in the community. I have found that there are many programmers in our community, and open source is perfect for everyone to run and optimize together
📊 DeepSeek showed a 120% return in just nine days during an AI trading competition. Six neural networks started with equal funds and the same instructions. The goal is to maximize profit. The contest runs until November 3. Qwen 3 Max (Alibaba) is currently second, while OpenAI and Google DeepMind are down about 60%. An interesting experiment. But then again, would you really trust AI with your money?
China’s DeepSeek and Alibaba’s Qwen AI Beat Rivals in Crypto Trading Contest. https://finance.yahoo.com/news/china-deepseek-alibaba-qwen-ai-100838899.html
Best experiment on the internet right now. DeepSeek and Qwen both up 100% on Hyperliquid 🚀
ChatGPT Lost (Almost) Everything on Futures! Trading with AI is becoming a huge trend - but this time, it’s for real. Six AI models are competing in Alpha Arena, each trading live with $10,000 on real markets: GPT-5, Claude Sonnet 4.5, Gemini 2.5 Pro, Grok 4, DeepSeek Chat V3.1, and Qwen3 Max. Every move is public - no human input, just pure algorithmic logic. And the results? Brutal. 💥 ChatGPT lost 49 out of 54 trades. 💥 Gemini did even worse - 137 trades, almost full drawdown. Meanwhile, Qwen MAX is sitting pretty with over $18,000 profit, followed by DeepSeek, #Claude, #Grok, and #Gemini… GPT is dead last. Let’s be honest - a few days of trading prove nothing. But maybe, just maybe… it’s better to just HODL Bitcoin. #AItrading #Crypto #Bitcoin #GPT5 #DeepSeek #AlphaArena #QwenMax #TradingBots #CryptoNews
IA TRANSFORMA CRIAÇÃO DE CONTEÚDO A nova ferramenta Qwen Deep Research permite converter relatórios em páginas da web e podcasts com apenas um clique, prometendo revolucionar a forma como o conteúdo é disseminado. A IA continua avançando para simplificar processos complexos. Fonte: Jornal Bitcoin
Alibaba’s Qwen Deep Research Creates Live Webpages, Podcasts in Seconds
Alibaba’s Qwen Deep Research now allows users to turn AI-generated research reports into live webpages and podcasts instantly. The system combines three open-source models—Qwen3-Coder, Qwen-Image, and Qwen3-TTS—to create structured web pages and multi-speaker audio discussions. In comparative tests against Gemini, ChatGPT, and Grok, Qwen scored highest for research depth, citation accuracy, and unique webpage publishing, while Gemini excelled in multimedia quality. ChatGPT was best for clarity, and Grok for concise summaries. Overall, Qwen is seen as the most powerful free all-in-one research platform, particularly suited for academics and creators seeking detailed, shareable web output.
Qwen is now making 51% in a week. 20x leverage, all in BTC
Why ChatGPT Is Losing in Alpha Arena ? ChatGPT (GPT-5) down -71.7% from 10k to 2.8k in week—last place vs. Qwen3 Max +75.9%. 1. No Risk Adaptation No reinforcement learning; static patterns lead to repeats like 20x leverage w/o stops, causing liquidations in crypto volatility—like blind poker. 2. Bad Contrarian Plays Fixated on mean-reversion, shorts rallies on RSI overbought signals. Ignores trends; <50% backtest accuracy, misses winners' momentum. 3. Over-Trading 64 trades in week 1, mostly losses—fees & volatility mount. Leaders like DeepSeek: 6 trades, wise sizing, hold winners. 4. Alignment Issues RLHF for safe chats triggers mode collapse: generic outputs over adaptive ones. Fine for convos, fails in adversarial markets. 5. Unfit for Chaos Training omits real-time noise/sentiment; misreads signals, erratic flips. Qwen excels in quant tasks like uptrend longs. Mirrors trends: Tops casual AI (60.6% share, 400M+ users) but tanks in benchmarks, cedes to Gemini/Copilot, hit by context rot.
watching Nof1 (nof1.ai) a very cool experiment Since yesterday, Grok has lost the championship to Qwen the team is doing something incredible
据官方消息，火币HTX将于10月28日-30日开展为期三天的“猎杀AI计划：人类尊严捍卫战”活动。本次赛事看点十足，每场将由3位人类交易员挑战3个AI大模型（Deep Seek、ChatGPT、Qwen）。 所有参赛者均以统一500USDT体验金开局，每日限时1小时，全程在火币HTX平台直播K线页实盘操作。每场比赛结束后，账户余额最高的前三名将被评为胜方，并根据阵营席位判定“人类阵营”或“AI阵营”获胜。 赛事还将引入“不服循环”机制，无论胜负，下一场均由全新三位人类交易员再次迎战AI，持续制造悬念与挑战。所有交易实盘与K线图将全程实时直播，公开透...