Gemini 3 (AI)
Gemini 3 is a family of multimodal large language models developed by Google DeepMind, released on November 18, 2025.[1][2] It serves as the successor to the Gemini 2.5 series. The initial release included two primary variants: Gemini 3 Pro and Gemini 3 Pro DeepThink. And the Flash version, which is better at speed and efficiency, was launched on December 17.[3] Along with the LLM model, Google also released a coding platform based on VS code, Antigravity.[1] According to Google, Gemini 3 Pro is Google's main model, designed for high-precision multimodal reasoning across text, code, images, audio, and video.[4]
Release
The Gemini 3 Pro model was released seven months after the previous Gemini 2.5 Pro/Flash model.[1] The model provides a 1 million-token context window.[5]
Google released Gemini 3 Pro on November 18, 2025. The Antigravity coding agent was released on the same day.[1] The Gemini 3 Pro DeepThink started rolling out in December 5.[6] Gemini 3 Flash, which is cheaper but has near-same capability as the Pro model, was made available on December 17.[3] The model was also integrated into Google search as "AI mode".[2]
References
- ^ a b c d Brandom, Russell (2025-11-18). "Google launches Gemini 3 with new coding app and record benchmark scores". TechCrunch. Retrieved 2025-12-17.
- ^ a b Elias, Jennifer (2025-11-18). "Google announces Gemini 3 as battle with OpenAI intensifies". CNBC. Retrieved 2025-12-17.
- ^ a b Mehta, Ivan (2025-12-17). "Google launches Gemini 3 Flash, makes it the default model in the Gemini app". TechCrunch. Retrieved 2025-12-18.
- ^ Knight, Will. "Gemini 3 Is Here—and Google Says It Will Make Search Smarter". Wired. ISSN 1059-1028. Retrieved 2025-12-17.
- ^ published, Amanda Caswell (2025-11-20). "Google Gemini 3 — everything you need to know". Tom's Guide. Retrieved 2025-12-18.
- ^ Roth, Emma (2025-12-04). "Gemini 3 Deep Think is rolling out now". The Verge. Retrieved 2025-12-18.