Gemini 4 Argon: Google is back among the top three labs in intelligence achieved (artificialanalysis.ai)

Google DeepMind has launched its latest AI model, Gemini 4 Argon, marking a significant return to the forefront of AI research, achieving a score of 53 on the Artificial Analysis Intelligence Index. This performance places it on par with OpenAI's GPT-6 Astra while offering a competitive price point of $1.99 per task, which is 60% lower than GPT-6 Astra’s cost. Notably, Gemini 4 Argon demonstrates remarkable advancements in reasoning and agentic functionality, a historical weakness for Google models, and the lowest hallucination rate among leading models, enhancing its reliability for practical applications. The model's architecture supports a vast context window of 1 million tokens and accommodates multiple modalities, including text, image, video, and speech input. It features improved performance on various benchmarks such as AutomationBench-AA and Terminal Bench 4, positioning it strongly within the AI landscape. Gemini 4 Argon's design not only boosts task performance while reducing costs, but its capacity for adaptive reasoning via the new Long Decode Continuation feature sets a new standard for future AI models. With its initial launch discounts, this model is poised to challenge existing competitors and redefine cost-efficiency in advanced AI applications.