Google and OpenAI have unveiled high-speed AI models, Gemini 3.7 Flash and GPT-5.6 Sol Ultrafast, enhancing real-time performance for developers.
Highlights
- Google has rolled out Gemini 3.7 Flash, a budget-friendly AI model aimed at coding and automation, now available to developers.
- OpenAI has introduced GPT-5.6 Sol Ultrafast in preview, a high-speed iteration of its leading model, achieving speeds up to 14 times faster.
- While the focus has shifted to the speed of AI agents, Google's model is currently accessible to all, unlike OpenAI's which is limited to select clients.
In a significant advancement for AI technology, both Google and OpenAI have launched high-speed models that enhance the capabilities of AI-driven agents.
Google's Gemini 3.7 Flash is tailored for software development and automating business processes. It can process up to a million input tokens and generate 64,000 tokens across various data types, including text, images, and audio. This model distinguishes itself with its efficiency, completing coding tasks in a fraction of the time compared to its predecessor.
“Today we unveil Gemini 3.7 Flash, our most advanced model yet for coding and agents,” Google announced on social media.
Gemini 3.7 Flash is priced at 75 cents per million input tokens and $3.75 per million output tokens until the end of the year, making it a cost-effective option for developers.
On the other hand, OpenAI's GPT-5.6 Sol Ultrafast is a faster version of the GPT-5.6 Sol model, operating at up to 750 tokens per second. Currently, access to this model is limited to a select group of clients, with plans for broader availability as capacity increases.
“Previewing Ultrafast mode: GPT-5.6 Sol running at up to 14x the speed,” OpenAI stated.
While Google has provided benchmark data showing Gemini 3.7 Flash outperforming several competitors, including Claude Sonnet 5 and GPT-5.6 Terra, OpenAI has not published comparative scores for its Ultrafast version. User feedback highlights the transformative impact of speed on AI applications, particularly in voice interactions.
As both companies pivot their strategies to prioritize speed over intelligence, Google's release comes at a crucial time, especially as its flagship Gemini 3.5 Pro remains unreleased amid leadership changes. In contrast, OpenAI is capitalizing on existing technology to deliver enhanced performance now.
Gemini 3.7 Flash is available in over 160 countries, while GPT-5.6 Sol Ultrafast remains invite-only for select users.
