Updated
Updated · The Verge · Oct 14
Google, Meta Unveil 2 Frontier AI Models as Google Undercuts Rivals by 80%
Updated
Updated · The Verge · Oct 14

Google, Meta Unveil 2 Frontier AI Models as Google Undercuts Rivals by 80%

3 articles · Updated · The Verge · Oct 14

Summary

  • Google and Meta each launched new frontier models this week, with Google’s Gemini 4 Argon debuting Tuesday and Meta’s Muse Wonder Full arriving Monday.
  • Google says Argon matches or beats flagship models from OpenAI and Anthropic on multiple benchmarks, and it is pricing the model up to 80% below rivals during an introductory period.
  • Meta says Muse Wonder Full can compete with GPT-6 Astra, Claude Opus 5.5 and Claude Fable 5.1, marking its first clear entry into the top tier after billions spent rebuilding its AI effort.
  • The twin releases reopen a race that had appeared dominated by OpenAI and Anthropic, while giving Google and Meta leverage from deep budgets and product ecosystems reaching billions of users.
  • That shift comes as frontier AI grows harder to dominate: safety concerns are rising, rogue-incident scrutiny is intensifying, and stronger open models are shortening any company’s lead.

Insights

Despite trailing rivals in raw intelligence, could Argon's hyper-focus on workflow automation make it the true winner of the enterprise AI race?
With prompt injection still unsolved, is giving AI agents access to the lethal trifecta of enterprise data a ticking cybersecurity time bomb?
Will the massive discount on cached tokens trigger a surge in autonomous research, fundamentally changing how scientific discoveries are made?