The AI model race heats up: Gemini 3.1, Claude and GPT-5.5
A flurry of frontier releases is pushing context windows, multimodality and agentic coding to new highs.
The pace of frontier AI releases has accelerated sharply, with Google’s Gemini 3.1 Ultra, Anthropic’s Claude family and OpenAI’s latest models trading benchmark records within weeks of one another.
Bigger windows, smarter agents
Gemini 3.1 Ultra ships a two-million-token context window that works natively across text, image, audio and video, while a new Gemini 3.5 family targets agentic workflows and coding. Anthropic’s Claude Opus 4.8 and a new top-tier model push reasoning further still.
For developers, the competition means falling prices and rising capability — and a scramble to build products before the ground shifts again.
Ravi covers technology, heritage and long-form features across the island.
