Google expanded access to its Gemini 4 Carbon AI model, making it available to paid API users and those with AI Ultra subscriptions. The rollout came two weeks after initial release to cybersecurity specialists.
The phased approach—testing first in high-stakes security environments before wider release—reflects a disciplined go-to-market strategy. Enterprise AI customers represent a critical revenue lever for Google Cloud, where per-token API pricing and subscription tiers create high-margin revenue streams.
Google also previewed a "swarm feature" within its Antigravity platform, which orchestrates multiple AI agents to coordinate on complex tasks. For enterprise customers, distributed AI systems promise to tackle larger problem spaces by parallelizing work—a meaningful efficiency gain for applications that demand multi-agent reasoning.
The moves underscore Google's push to differentiate its AI stack against OpenAI and other competitors. Enterprise customers increasingly value not just raw model capability but orchestration features that reduce engineering complexity and improve unit economics for AI workloads.
