List of LLM, Image, Video, and Audio Models
Last Surveyed: Feb 2, 2026
Gemini 3.1 Pro suffers from severe 'context amnesia'. While leaked v3.5 shows God-tier reasoning, the current public 3.1 is failing at basic recall tasks.
Persistent OAuth failures and account lockouts have rendered Claude unusable for production. Many users are migrating back to ChatGPT or Cursor.
New 'Thinking' models now verbalize their plans before execution. It catches edge cases but doesn't fix the underlying 'Laziness' when writing large files.
Major services are failing at the infrastructure level. 'Amnesia' bugs and authentication lockouts are widespread. Do not trust AI with raw data without offline backups.