Gemini 3 Pro
Overview
Gemini 3 Pro is a large language model developed by Google DeepMind, released in November 2025. It is designed to handle complex reasoning tasks across multiple modalities, including text, images, audio, and video. The model features a 1 million-token context window, allowing it to process extensive inputs effectively. Its architecture is based on a Transformer-based Mixture-of-Experts model, enabling efficient handling of diverse data types. Gemini 3 Pro has demonstrated strong performance in various benchmarks. It achieved a score of 91.8% on the MMLU (Massive Multitask Language Understanding) benchmark, indicating a high level of proficiency across a wide range of academic subjects. In coding tasks, it scored 76.2% on the SWE-bench-Verified benchmark, reflecting its capability in real-world programming scenarios. Additionally, it achieved a 95.0% score on the AIME-2025 (American Invitational Mathematics Examination 2025), showcasing its mathematical reasoning abilities. These results position Gemini 3 Pro as a competitive model in the AI landscape, particularly for applications requiring advanced reasoning and multimodal understanding. As of June 2026, Gemini 3 Pro is available through Google's API, with pricing set at $2.50 per million input tokens and $20.00 per million output tokens. Its release has been followed by the introduction of Gemini 3.1 Pro in February 2026, which offers enhanced reasoning capabilities and improved performance across various benchmarks. Users are encouraged to consider Gemini 3.1 Pro for tasks requiring advanced reasoning and multimodal processing.