08:14 Benchmarks2 months ago Penguin-VL in 2B and 8B: Worst Vision AI Model Ever: Full Local Testing Fahd Mirza puts Tencent's newly released Penguin-VL vision-language models — available in 2B and 8B parameter sizes — through a serie... 0 comments 765 views
09:51 Research & Benchmarks2 months ago Tri-21B Think: Breaking US-China Duopoly in AI Dominance: Run Locally Fahd Mirza takes a hands-on look at Tri-21B Think, an open-weight reasoning model from Trillion Labs, a South Korean AI startup. The... 0 comments 4.8K views
12:24 Research & Benchmarks3 months ago OpenFang: Hype or Real? Open Surgery of Yet Another Claw Variant Fahd Mirza delivers an unusually candid teardown of OpenFang, a self-styled \"agent operating system\" written in Rust that claims to... 0 comments 7.8K views
08:07 Research & Benchmarks3 months ago NeuTTS Nano Multilingual: Great Idea, Disappointing Execution Fahd Mirza tests NeuTTS Nano's newly released multilingual voice cloning models — French, Spanish, and German variants — alongside th... 0 comments 1K views
11:06 Tutorials3 months ago Qwen3.5-122B-A10B: The Ultimate AI Party: Run Locally with Style Fahd Mirza demonstrates running the Qwen3.5-122B-A10B mixture-of-experts model on a single Nvidia H100 80GB GPU using llama.cpp — a f... 0 comments 4.3K views
09:40 Tutorials3 months ago NullClaw + Ollama Local Setup with Telegram Fahd Mirza walks through installing and configuring NullClaw, a fully autonomous AI assistant written in Zig — a low-level systems pr... 0 comments 14.4K views