Is OpenAI loosing the AI race?

Here’s how I see it:

• OpenAI is still the hype engine, but not obviously the value capture engine. ChatGPT was the tool that made LLMs mainstream in late 2022. People think about it like 'iPhone' but maybe it's just a Blackberry or a Nokia. Here is why:
  • Google just launched Gemini 3, plugged straight into Search and a new agent-first coding IDE (Antigravity).
  • Benchmarks show Gemini 3 Pro slightly edging out GPT-5.1 on some reasoning benchmarks, while Google uses it to defend its core money-printer (Search). blog.google+2The Verge+2
  • Meanwhile OpenAI has GPT-5.1 + o3 + Sora 2, but a lot of the actual revenue looks like it flows through Microsoft Copilot and partners, not purely OpenAI-branded products. The GitHub Blog+3OpenAI+3OpenAI+3
  • If Google and OpenAI launch the exact same products, Google still win on the long run. The competitive edge becomes the data that Google has on the end user.
• OpenAI built the general tool; others are nailing specific use cases. OpenAI is basically “AI for everyone” (horizontal, general-purpose). But in verticals:
  • Google is turning Gemini 3 into a thought partner inside Search and a full IDE with agents (Antigravity). blog.google+1
  • The Browser Company, Perplexity, etc. are pushing AI-native browsers and search UIs as their only job. OpenAI’s own Atlas browser exists, but it’s one player in a crowded “AI browser” space with no strong teams.  
  • Chinese labs are shipping agentic features like Kimi’s “OK Computer” (build full sites/slides from prompts) and DeepSeek-style reasoning agents at aggressive pricing.
• The competitive field is way more crowded than “OpenAI vs the world”. It’s not “OpenAI and maybe LLaMA” anymore. Here is what is happening now:
  • Gemini 3, Claude, DeepSeek, Kimi, open LLaMA/Qwen variants…
  • DeepSeek’s R1 openly claims o1-level reasoning at a fraction of the cost, and its low-price APIs triggered an AI price war in China and spooked global markets. 
  • Moonshot’s Kimi K2 is open-weight and ridiculously cheap per token compared to GPT-4/5-tier models. 
• OpenAI is carrying a disproportionate share of the blame and legal risk. Any time something goes wrong with AI, “ChatGPT” is the headline, even when it’s not actually the tool used. OpenAI is: Other companies (Google, Meta, Anthropic…) are also getting sued and criticized, but OpenAI is the symbol everyone points at. That slows them down:
  • Being sued over copyright by news orgs, authors and music rights groups (NYT, GEMA, Ziff Davis, etc.). 
  • At the center of debates about AI psychosis, suicide risk, and mental health, with OpenAI itself now admitting hundreds of thousands of users weekly show signs of serious crises in chat logs. 

So my feeling right now is:

OpenAI is still one of the leaders on quality and adoption, but no longer the obvious winner. They are focusing mostly now on B2B (recent intuit deal, Microsoft partnership…)
The real “AI race” is turning into a price + integration + ecosystem game, not just “who has the fanciest demo”.

TL;DR:

• OpenAI kicked off the boom with ChatGPT, but Google, DeepSeek, Kimi, Claude, etc. are now matching or beating it on reasoning, price, or integration.
• Google has the unfair advantage of Search + user data + product distribution: if it ships the same features as OpenAI, it probably wins over time.
• Chinese labs are redefining the game with o1-level reasoning at a fraction of the cost, making this a price + ecosystem war, not a “cool demo” war.
• OpenAI still leads on quality and adoption, but it’s carrying most of the blame, lawsuits, and regulatory heat, while shifting more into B2B (Copilot, Intuit, enterprise deals).