The launch of Google’s Gemini 3 has reportedly triggered an OpenAI “code red,” escalating the fierce rivalry in the AI sector. Two years after ChatGPT disrupted the market, Google’s powerful multimodal model is rapidly gaining on it in user numbers, forcing OpenAI to defend its leadership against new competition, pricing pressures, and shifting enterprise buying habits.
Gemini 3 Raises the Temperature
OpenAI declared a “code red” after benchmarks revealed Google’s Gemini 3 surpassed its own models on key tasks like image reasoning. This, combined with Gemini’s rapid user growth to 650 million, prompted OpenAI to halt secondary projects and refocus all resources on enhancing ChatGPT’s core capabilities.
Google prioritized a rapid launch, and early benchmarks indicate Gemini 3 outperforms GPT-5 in critical areas like image reasoning and long-context tasks. This performance earned praise from industry leaders like Salesforce CEO Marc Benioff. In response, a leaked memo confirmed Sam Altman initiated a “code red” on December 1, halting non-essential projects to concentrate all efforts on improving ChatGPT OpenAI Declares Code Red as Google’s Gemini 3 Advances. The threat is amplified by user growth; Axios reports Google’s user base surged 44% in one quarter to 650 million users Google Gemini 3 Pro levels up to OpenAI’s ChatGPT. This growth is driven by a generous free tier, developer discounts on Vertex AI, and a $19.99 Gemini Advanced subscription bundled with Google One.
OpenAI’s Strategic Response: Product Velocity and User Experience
OpenAI’s primary response is to accelerate product development. While Sam Altman has promised a new model to outperform Gemini 3, the internal focus has shifted from raw performance to a frictionless user experience. Recent updates, such as group chats for up to 20 users and memory features that personalize interactions, are designed to increase user stickiness and create switching costs.
The two companies also differ in their monetization strategies. OpenAI relies heavily on direct subscriptions, whereas Google diversifies its revenue through ads, Workspace, and API usage. This allows Google to absorb lower per-user margins, putting pressure on OpenAI to justify its $20/month subscription with tangible feature enhancements instead of price reductions.
The Enterprise Shift to Multi-Model AI Strategies
In the enterprise sector, CIOs are diversifying their AI portfolios instead of relying on a single provider. An a16z survey of 100 global firms reveals that 37% now use five or more foundation models in production How 100 Enterprise CIOs Are Building and Buying Gen AI in 2025. This multi-model approach enables companies to avoid vendor lock-in and select the optimal model for each task based on performance and cost. Platforms from vendors like Liminal facilitate this strategy by providing unified governance and security across different models Multi-Model AI Platforms vs. Single Provider. As one CIO noted, with leading models reaching parity for many tasks, “procurement now optimizes for cost and compliance.” This pragmatic view puts further pressure on pricing as OpenAI and Google compete for enterprise contracts.
- Quick-scan factors shaping 2025 purchasing decisions:
- Context-window length for document workflows
- Token pricing schedules at scale
- Built-in governance features (DLP, audit logs)
- Road-map transparency and support SLAs
Despite the pressure, OpenAI’s significant brand loyalty remains a key asset. With ChatGPT reportedly powering 70% of global AI-assisted activities, its user base forms a substantial competitive moat. However, the history of tech shows that such advantages can erode quickly in the face of disruptive innovation. The next few quarters will be critical in determining if OpenAI’s strategy of rapid innovation and user-centric features is enough to counter Google’s multifaceted assault with Gemini.
What triggered OpenAI’s “code red” and how is the company reacting?
Sam Altman told staff on 1 December that ChatGPT is at a “very critical juncture” and ordered every team to drop side projects and focus only on improving ChatGPT.
The move came after Google’s Gemini 3 (released 18 November) beat OpenAI’s internal GPT-5 benchmarks and helped Gemini’s monthly active users jump from 450 M in July to 650 M in October while maintaining higher user-satisfaction scores, especially for image tasks.
OpenAI has postponed its planned ad network, shopping agents and “Pulse” report generator so engineering and product resources can ship ChatGPT enhancements faster.
How might the rivalry change the way AI assistants make money?
ChatGPT today earns almost all of its revenue from $20-per-month Plus/Pro subscriptions and API fees; Gemini spreads income across $19.99 Google-One bundles, enterprise Workspace licences, cloud-API tokens, search-ad boosts and Pixel-device sales.
OpenAI has already tested ad formats inside ChatGPT, but the “code red” could accelerate a dual model – keep premium tiers ad-free while inserting sponsored prompts or branded plugins for free users.
For enterprises, expect sharper price competition: Gemini already undercuts GPT-4-level pricing on long-context workloads, and OpenAI is likely to respond with volume discounts and new mid-tier plans in 2025.
Should companies hedge their AI bets instead of picking one vendor?
Yes. A 2025 survey of 100 CIOs shows 37 % of large firms now run five or more models in production, up from 29 % last year.
Multi-model setups let teams match each workload to the best-priced, best-performing model and avoid single-vendor outages, sudden price hikes or roadmap changes.
Enterprise-grade gateways add unified audit logs, data-loss-prevention rules and role-based access, so adding or swapping models does not create new compliance headaches.
What features are ChatGPT and Gemini rolling out to keep users loyal?
OpenAI just opened 20-person group chats to every plan tier and is upgrading memory-based personalization that remembers user style, code libraries and project history across sessions.
Gemini 3 ships with a 1-million-token context window, deeper Workspace sidebar ( Docs, Sheets, Meet) and on-device Nano for Pixel 8/9 phones that works offline.
Both firms are racing to release agent-style tools – OpenAI delayed its shopping bot but is polishing in-browser task execution, while Gemini touts trip-booking and form-filling agents already in private alpha for Google One subscribers.
Which model leads on brand reach and switching costs today?
ChatGPT still accounts for ~70 % of global AI-assistant traffic and 10 % of search-related queries, giving it strong habit-forming power and the largest plug-in ecosystem.
Yet Gemini’s 650 M monthly users plus default placement in Google Search, Android and Chrome create massive passive exposure that ChatGPT cannot match without paying for placement.
Switching costs are rising on both sides: memory, shared projects and paid API integrations make it painful to leave ChatGPT, while Gmail, Docs and Drive context embedded in Gemini replies make Google’s suite stickier for workplace teams.
