{"slug": "gemini-3-5-pro-s-delay-just-cost-alphabet-nearly-200-billion-in-a-day", "title": "Gemini 3.5 Pro's Delay Just Cost Alphabet Nearly $200 Billion in a Day", "summary": "Alphabet lost nearly $200 billion in market value on July 16 after Bloomberg reported that Google's Gemini 3.5 Pro AI model is months behind schedule due to weak coding performance. The stock fell 4.4% as investors reacted to the delay in the flagship model, which is critical in the competitive AI coding market.", "body_md": "*Alphabet shed nearly $200 billion in market value in a single session on July 16 after Bloomberg reported that Gemini 3.5 Pro, Google's flagship AI model, is months behind its own June deadline because of weak coding results.*\n\nThe stock fell 4.4% that day, and the drop had nothing to do with earnings or a lawsuit. It came down to a missed deadline. Google introduced Gemini 3.5 Pro in May, calling it the most capable model the company had ever built, and told investors a broad rollout would follow in June. June came and went. All Google would say was that it's \"currently testing 3.5 Pro, an upgraded Flash model, and other models with partners.\" No launch date. No explanation, at least not from the company itself.\n\nBloomberg's reporting filled that gap. According to Bloomberg, Google retrained Gemini late last month specifically to sharpen its coding performance, and the results went the wrong way: older checkpoints of the model actually outperformed the newer ones on internal coding evaluations. That's the kind of setback that doesn't get fixed with a press release. Ten current and former Google employees told Bloomberg they're frustrated, worried the company is ceding ground in coding, the single most contested battleground in AI right now, to Anthropic and OpenAI.\n\n## The Price Tag\n\nHere's the thing about that $200 billion figure. It's larger than Alphabet's entire annual AI infrastructure budget. The company has guided investors to expect $175 billion to $185 billion in capital expenditures for 2026, according to CNBC's coverage of Alphabet's February earnings call. That's more than double what it spent building AI infrastructure the year before. Markets erased more value in one afternoon than Google plans to spend building data centers and buying chips for the entire year. That's not how you'd expect a single product delay to be priced. Unless investors have started treating model release timing as a leading indicator of who actually wins this race.\n\nCoding has become the metric that matters most in the current AI cycle, more than chatbot fluency or trivia scores. Enterprises buy AI tools to write and review code, not to pass philosophy exams. Every major lab knows it. That's why Google's own coding shortfall stings more than a generic delay would. The company built its pitch around Gemini 3.5 Pro being its most powerful release yet, and then quietly confirmed, months past deadline, that it still isn't ready.\n\n## Rivals Are Closing In\n\nThe timing makes it worse. On July 16, the same day Bloomberg's report landed, Moonshot AI released Kimi K3, a 2.8 trillion parameter open weight model that took the top spot on the Frontend Code Arena leaderboard, jumping seventeen places past its predecessor and ahead of Anthropic's Claude Fable 5 on that specific benchmark. Kimi K3 still trails Fable 5 and GPT-5.6 on some broader measures. But the direction of travel is unmistakable. Chinese labs are closing the gap on the exact capability where Google just admitted it's stumbling.\n\nAnthropic and OpenAI haven't been sitting still either. Both have shipped newer models over recent months that reset expectations for coding and agentic tasks, part of the wider run of releases that outlets have tied to Gemini's slipping edge. Every one of those releases raises the bar Gemini 3.5 Pro now has to clear. And every month of delay gives rivals more runway to lock in developers, enterprise contracts, and the benchmark headlines that decide which model gets tried first.\n\nNone of this means Alphabet is in real trouble. The company still has search, YouTube, cloud, and a balance sheet most of Silicon Valley would kill for. But the market reaction tells you something specific. Investors want the ship date to hold, and when it doesn't, they're now willing to price that failure at nearly the size of an annual infrastructure budget in a single trading session. Google still hasn't given Gemini 3.5 Pro a new release date. Until it does, that uncertainty is the story.\n\n**Also read:** [OpenAI's GPT-5.6 Sol can already find browser exploits, just not finish them](https://startupfortune.com/openais-gpt-56-sol-can-already-find-browser-exploits-just-not-finish-them/) • [SpaceX is reportedly chasing a Pentagon AI compute deal](https://startupfortune.com/spacex-is-reportedly-chasing-a-pentagon-ai-compute-deal/) • [AWS billing dashboard told a customer who spent 19 cents that they owed 2.5 billion dollars](https://startupfortune.com/aws-billing-dashboard-told-a-customer-who-spent-19-cents-that-they-owed-25-billion-dollars/)", "url": "https://wpnews.pro/news/gemini-3-5-pro-s-delay-just-cost-alphabet-nearly-200-billion-in-a-day", "canonical_source": "https://startupfortune.com/gemini-35-pros-delay-just-cost-alphabet-nearly-200-billion-in-a-day/", "published_at": "2026-07-18 02:53:26+00:00", "updated_at": "2026-07-18 03:04:52.630797+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "ai-products", "ai-ethics", "ai-policy"], "entities": ["Alphabet", "Google", "Gemini 3.5 Pro", "Bloomberg", "Anthropic", "OpenAI", "Moonshot AI", "Kimi K3"], "alternates": {"html": "https://wpnews.pro/news/gemini-3-5-pro-s-delay-just-cost-alphabet-nearly-200-billion-in-a-day", "markdown": "https://wpnews.pro/news/gemini-3-5-pro-s-delay-just-cost-alphabet-nearly-200-billion-in-a-day.md", "text": "https://wpnews.pro/news/gemini-3-5-pro-s-delay-just-cost-alphabet-nearly-200-billion-in-a-day.txt", "jsonld": "https://wpnews.pro/news/gemini-3-5-pro-s-delay-just-cost-alphabet-nearly-200-billion-in-a-day.jsonld"}}