Google has launched a new model called Gemini 2.0 Flash Thinking that offers advanced reasoning capabilities and transparency in AI problem-solving.
The model can solve complex problems at lightning speeds and display its internal planning process to developers.
Gemini 2.0 Flash Thinking competes with OpenAI’s o1 model which is known for great reasoning capabilities similar to PhD students in various sciences.
Google has launched Gemini 2.0 Flash to support multimodal inputs like images, video, and audio for TTS (text-to-speech).
A few benchmarks suggest that o1 is the most powerful AI model beating even the benchmark-claude-3.5-sonnet in coding tasks.
The only drawback of o1 is that it takes longer than other models to complete tasks.
Google is ahead of OpenAI, which is now playing catchup in the AGI race.
While OpenAI has been making announcements in their 12 Days of OpenAI, Google introduced several innovations.
Google has previously launched quantum chip Willow, Gemini 2, the 3D world model Genie 2, Veo 2, Project Astra as a universal agent, Project Mariner, Google Deep Research, and Android XR.
Gemini 2.0 Flash Thinking builds on Google's Gemini series which competes with OpenAI's o1 model.