TSTidiane Stanoin4sapi.hashnode.dev·5d ago · 8 min readWhy Open Models Start From Qwen: AI Base Models ExplainedAbstract The open‑source large‑model ecosystem is undergoing a clear shift. Numerous newly‑released open‑source models are not built from scratch; instead, they are fine‑tuned, continuously pre‑traine00
Ccoffeeinxagent.hashnode.dev·Aug 29 · 5 min readI Enabled Wireless Debugging. xAgent and a 27B Quantized Model Did the RestI started with a simple question: if I gave xAgent a test server and a phone, could it route the phone's traffic through the server? It ended up installing mitmproxy and Frida, connecting ADB and the 00
NNickinblog.gpus.market·Aug 14 · 9 min readDeploying Qwen3.8-2.4T-A95B with vLLM: Verified GPU Pods, Quants, and Serving RecipesQwen3.8-2.4T-A95B is a 2.4-trillion-parameter Mixture-of-Experts model with roughly 95B parameters active for each token. If you're planning to self-host it, the first thing to know is that this is a 21T
MAMuhammad Azlaan Zubairinblog.mdazlaanzubair.com·Aug 8 · 16 min readClaude’s Intelligence Isn’t Worth $20 to Me AnymoreI like Claude. That is precisely why I have a problem with Claude Pro. Claude can write exceptionally well. Its coding models are genuinely competitive. Claude Code is one of the strongest arguments 24TTD
DSDarsh Shahinfreecodecamp.org·Aug 6 · 16 min readHow to Customize an LLM for AI Agents using SFT and QLoRAIn this tutorial, I’ll show you how to fine-tune a large language model for use in AI agents using supervised fine-tuning with QLoRA. This lets us customize a pre-trained model so it behaves the way w10
AGAditi Guptainhonestaireview.hashnode.dev·Aug 4 · 8 min readQwen3.8-Max Beat ClaudeAlibaba has released Qwen3.8-Max, its largest and most capable AI model to date. The headline specifications are absurd: 2.4 trillion total parameters, 95 billion activated per token, multimodal input00
MSManu Shuklainecorpit.hashnode.dev·Aug 4 · 16 min readQwen3.8-Max went GA on 3 August 2026 at $2 and $6 per million tokens, and it is more expensive than Qwen3.7-Max is todayQwen3.8-Max went GA on 3 August 2026 at \(2 and \)6 per million tokens, and it is more expensive than Qwen3.7-Max is today Summary. Alibaba's Qwen team made Qwen3.8-Max generally available on 3 August00
MSManu Shuklainecorpit.hashnode.dev·Jul 31 · 15 min readQwen3.7 Flash vs Gemini 3.5 Flash-Lite: the 2026 cost math for 1M-context vision agentsQwen3.7 Flash vs Gemini 3.5 Flash-Lite: the 2026 cost math for 1M-context vision agents Summary. Alibaba released Qwen3.7 Flash on 27 July 2026 at \(0.03 per million input tokens and \)0.13 per millio00
DSDarsh Shahinfreecodecamp.org·Jul 24 · 11 min readHow to Use Prompt Engineering and Context Engineering for AI AgentsIn this tutorial, I’ll show you how prompt engineering and context engineering can improve an AI agent's performance. We’ll build a simple local agent, start with a baseline input, then improve it wit10
YFYuki Furutainblog.yukifuruta.com·Jul 22 · 24 min readRunning Edge LLMs on a Raspberry PiWhat if a Raspberry Pi could generate text entirely on its own, without sending prompts to a cloud-based AI service? To find out, I installed several compact large language models on a Raspberry Pi 5 00
