Google Rolls Out Gemini 4 Argon Amid Internal Debate Over Performance

Google's latest artificial intelligence model, Gemini 4 Argon, has been rolled out to the market despite internal debate over its performance on practical coding tasks. The model has shown strong results on industry benchmarks but some employees have raised concerns that it may not be as effective in real-world engineering work.

According to people familiar with the project, models from OpenAI and Anthropic are improving faster than Gemini 4 Argon. However, Google disputed these claims, pointing to internal testing and the remarks of Google DeepMind chief Koray Kavukcuoglu, who expressed confidence in the team's progress.

The release of Gemini 4 Argon follows months of work by Google to advance its AI lineup. The company had previously planned to release Gemini 3.5 Pro in June but abandoned that version as development shifted towards the newer model.