Gemini is quietly losing ground to ChatGPT and Claude

Three datasets now show Google's Gemini losing users to ChatGPT and Claude.

1 min
Gemini is quietly losing ground to ChatGPT and Claude

Three independent datasets now point the same way: people are using Google's Gemini less, not more.

AI-text detection firm Pangram measured model market share from the writing that users submitted for checking. OpenAI has held above 50 percent every month since Pangram began tracking. Anthropic climbed from 4.3 percent to 14.9 percent, mostly in technical and scientific work. Gemini fell from 12 percent to 1.9 percent in July 2026, a drop Pangram calls a "collapse." OpenRouter's usage rankings show the same direction.

Web analytics firm Similarweb adds a second view. Google's own website share slipped from 27 percent to 26.8 percent last month, though that is still up from 9.4 percent a year earlier. Google says Gemini's app has one billion monthly users, but monthly active users is a weak signal of real engagement.

The numbers help explain a restless few weeks at Google DeepMind, where several senior leaders have departed, including Demis Hassabis. None of the datasets is perfect: Pangram mostly captures writing, OpenRouter skews toward open-weight models, and Similarweb misses mobile apps. Together, though, they describe a model that is being used less even as Google pushes it harder.

AI market share comparison: one bar collapsing toward zero

Sources:

Written by

More to read

  • Anthropic Demonstrates Autonomous De Novo Protein Design and Chemical Analysis with Claude

    Anthropic Demonstrates Autonomous De Novo Protein Design and Chemical Analysis with Claude Anthropic has published experimental results demonstrating Claude's ability to autonomously design de novo protein binders with physical wet-lab validation and automate complex analytical chemistry workflows. The findings show frontier LLMs acting as autonomous agents across computational biology and molecular characterization pipelines. In the primary experiment, Anthropic evaluated Claude Mythos Previe

    1 min
  • Cerebras Unveils CS-4 Rack-Scale System Powered by Three WSE-3 Turbo Chips and Nexus Architecture

    Cerebras Unveils CS-4 Rack-Scale System Powered by Three WSE-3 Turbo Chips and Nexus Architecture Cerebras Systems has announced the CS-4, a rack-scale AI accelerator system designed around three of its next-generation Wafer Scale Engine 3 Turbo (WSE-3 Turbo) chips and a modular hardware architecture dubbed Nexus. Cerebras confirmed that initial customer shipments for the CS-4 are scheduled to begin in the current quarter. The new system marks a structural shift from Cerebras's single-wafer CS

    1 min
  • AI FinOps: Cutting LLM Inference Costs by 30-60% Through Model Tiering, Caching, and GPU Optimization

    AI FinOps: Cutting LLM Inference Costs by 30-60% Through Model Tiering, Caching, and GPU Optimization Inference costs have become the second-largest line item in enterprise AI budgets, trailing only talent spend according to RapidData's State of Enterprise AI 2026. This shift represents a fundamental inversion from the 2021-2023 era when training dominated AI expenditure. The compounding nature of serving costs—accumulating every hour as long as users hit the API—means that even modest producti

    1 min