Google launches Gemini 2.5 AI model amid ChatGPT’s 43% market lead

Google has introduced Gemini 2.5 Pro Experimental, a multimodal AI reasoning model designed to enhance coding, scientific, and mathematical tasks through deliberate “thinking” processes before generating responses.

The model, available via Google AI Studio and the Gemini app for $20/month subscribers, claims superiority over prior iterations and select competitors in benchmarks like LMArena’s WebDev leaderboard, where it ranks second with a score of 1,267.70.

Anthropic’s Claude 3.7 Sonnet retains the top spot in web development challenges with a score of 1,354.01, while DeepSeek and ChatGPT trail behind.

Google emphasises Gemini 2.5’s ability to support context-aware agents and process multimodal data (text, audio, images, code) with a 1 million-token context window, expandable to 2 million tokens.

OpenAI’s ChatGPT maintains dominance in the broader AI tools market, holding 43% market share as of February 2025, per aitools.xyz data.

DeepSeek ranks third with 6.58%, driven by a 195% growth rate and 792 million monthly visits.

While Google positions Gemini 2.5 as a step toward autonomous AI agents, industry observers note the model’s practicality in coding and integration over competitors like Claude and GPT-40.

The release underscores ongoing competition in AI reasoning models, with Google, Anthropic, and xAI prioritising accuracy and adaptability over raw speed.

Developers and enterprises increasingly evaluate these tools for reliability in specialised tasks, though ChatGPT’s entrenched user base complicates market shifts.