2 min read

OpenAI’s New Reasoning Model Outperforms Gemini 3.0 as Apple Drops Open-Weights Video Tool *StarFlow-V*

New AI Models and Benchmarks

OpenAI to Release New Reasoning Model, Outperforming Google’s Gemini 3 in Internal Tests

  • OpenAI is reportedly launching a new reasoning model next week, which internal evaluations suggest surpasses Google’s Gemini 3.0 in performance. The announcement follows delays in benchmarking for Gemini 3.0 and Anthropic’s Opus 4.5, while OpenAI’s GPT-5.1 Codex Max was evaluated rapidly.

Apple Releases Open-Weights Video Model StarFlow-V


UncensorBench: New Benchmark Exposes Flaws in AI Model "Abliteration"


AI Development Tools and Frameworks

Unsloth Enables 500K+ Context Length Fine-Tuning on a Single GPU


Multi-Agent Story Engine Built with Mistral Agents


AI Product Launches and Tools

Free Chrome Extension Enhances AI Prompts Across Multiple Platforms


AI Industry Partnerships and Enterprise Adoption

Accenture and OpenAI Partner to Integrate ChatGPT Enterprise


Mistral AI Partners with HSBC for Multi-Year AI Banking Solutions