Claude Opus 4.6: Is It Getting Dumber?
Developers report declining performance in Claude Opus 4.6, sparking debates on model drift and reliability.
📂 LLM News
Latest articles in LLM News
DeepSeek R1: Open-Source AI Challenges OpenAI
DeepSeek releases R1 model, offering open-source reasoning capabilities that rival top proprietary models at a fraction …
DevRouter.ai Offers $1.40 Credit for 99.99% Stable AI Routing
New AI proxy DevRouter.ai offers $1.40 credit to test its 99.99% stable routing system, featuring dynamic pricing and ac…
Redis Creator Runs DeepSeek Locally on Mac
Salvatore Sanfilippo's new ds4 engine lets users run DeepSeek V4 Flash locally on Apple Silicon, eliminating API costs.
Claude vs OpenAI: Can New Model Challenge Opus?
Developers debate if the latest Claude model can surpass OpenAI's Opus in coding rigor and logical consistency.
Avemujica Launches GPT-5.5 API Proxy with 10U Bonus
New Avemujica API proxy offers access to GPT-5.5 and Image-2 models. Register now for a 10U credit bonus.
OpenAI's Jakub Pachocki: Hiding Chain-of-Thought to Prevent Human Bias
OpenAI Chief Scientist Jakub Pachocki reveals why AI models hide reasoning steps, emphasizing the need for autonomous en…
Mistral Large 3 Rivals GPT-4o at Half the Cost
Mistral AI launches Large 3, a frontier model matching GPT-4o performance across key benchmarks while charging roughly 5…
Gemini 2.5 Ultra Tops Math Benchmarks
Google DeepMind's Gemini 2.5 Ultra achieves record scores on major mathematical reasoning benchmarks, surpassing GPT-4o …
Claude 4 Sets New Bar for Graduate-Level AI Reasoning
Anthropic's Claude 4 achieves state-of-the-art results on graduate-level reasoning benchmarks, surpassing GPT-4o and Gem…
OpenAI Launches GPT-5 Turbo with Multimodal Reasoning
OpenAI unveils GPT-5 Turbo, featuring native multimodal reasoning across text, image, audio, and video inputs.
OpenAI Launches GPT-6 Preview API
OpenAI unveils GPT-6 preview with native multimodal reasoning, offering developers unified text, image, audio, and video…