# Google Should Open Source Gemini. All of It > Source: > Published: 2026-07-09 17:31:30+00:00 Google gave the world the transformer. China is now giving the world the models. That is the problem. In 2017, Google researchers published “Attention Is All You Need,” the paper that introduced the Transformer architecture and became the foundation for modern large language models. ChatGPT, Claude, Gemini, Qwen, DeepSeek, GLM, Kimi all live in the world Google helped create. But Google no longer controls the direction of that world. It invented the substrate, then watched the AI industry turn that gift into companies, products, developer ecosystems, and trillion-dollar expectations. Now the same pattern is repeating with open-weight models. Only this time, the center of gravity is not Mountain View. It is moving to China. And that’s why Google should open source Gemini. Like, today. Not Gemma. Not a small edge model. Not a polite developer demo that runs on a laptop and lets Google say it participates in open AI. The actual flagship weights. The model family that would instantly become the American open-weight default if Google had the nerve to release it. This is not a charity move either, it’s a real strategy. Google is already fighting a brutal (and losing) closed-model war against OpenAI and Anthropic. In that arena, Gemini has to win on product polish, benchmarks, distribution, developer love, enterprise trust, pricing, speed, reasoning, coding, agents, and brand perception at the same time. That is a hard war to win when the market increasingly sees frontier AI as a rotating leaderboard. The closed API race is also unforgiving. If Gemini slips, everyone notices. At Google I/O 2026, Sundar Pichai said Gemini 3.5 Pro was coming “next month.” Google later reportedly pushed the release target into July while it gathered feedback and tuned the model. That delay is not proof that Google is incompetent. It may be the opposite. It suggests a team refusing to ship a model before it is ready. But that is exactly my point. Closed frontier AI forces Google to play the same product-release game as everyone else. Every missed window, every benchmark, and every delay becomes a story and evidence that Google is somehow falling behind in the very field it helped invent. Open weights would change the game. As a closed product, Gemini is one more model in a crowded API market. As an open-weight release, Gemini becomes infrastructure. Developers do not merely test it. They build around it. Enterprises do not merely call it. They deploy it, fine-tune it, audit it, host it, wrap it, and make it part of their internal stack. That is how defaults are created. Google knows this better than anyone. Android, Chromium, Kubernetes, and TensorFlow are some of Google’s most important wins that came from making technology too useful and too available to ignore. The transformer paper was probably the most valuable developer marketing Google ever produced, and it did not even look like marketing. It looked like research. That was the magic. They should do it again, but on purpose. The case for this is even stronger because the open-weight market is not waiting for America to organize itself. Z.ai released GLM-5.2 in June 2026, and Artificial Analysis ranked it as the leading open-weight model on its Intelligence Index shortly after launch. Z.ai described the model as built for long-horizon tasks, while outside observers highlighted its MIT-licensed open-weight release and serious coding performance. Reuters has reported that Chinese open-source models are being widely adopted globally because of their cost and technical strength, and that Beijing is now considering restrictions around access to some AI technologies. That should terrify anyone who cares about American AI leadership. The West spent years assuming open models were the low-end of the market. Then Chinese labs used them to win developer mindshare, enterprise experimentation, and global distribution. The economics are obvious. Most enterprise AI work is not Nobel-level reasoning. It is document review, extraction, search, coding assistance, email triage, workflow automation, customer support, analytics, summarization, and internal agents. For that work, “good enough and cheap enough” beats “slightly better and far more expensive” all day. That is where open weights become a winner. Once a company standardizes on a model family, the model becomes more than a tool. It becomes a platform decision. Fine-tunes accumulate. Eval harnesses get built. Internal prompts get written. Security reviews get approved. Engineers become familiar with its quirks. Switching costs appear. If American companies standardize on Qwen, DeepSeek, Kimi, GLM, or whatever Chinese model wins the next open leaderboard, that is not just a pricing story. It is a platform story. The default AI stack for global enterprises starts drifting toward Chinese model families because American frontier labs kept their best systems locked behind API meters. Google is the one company that can stop that without waiting for Washington, OpenAI, Meta, or anyone else. OpenAI did release gpt-oss, its first open-weight model family since GPT-2, with 120B and 20B parameter versions under Apache 2.0. But OpenAI’s own materials describe those models as open reasoning models designed to run locally and in data centers, not as the company’s flagship frontier line. They are important, but they are not OpenAI giving away the crown jewels. Google’s Gemma family is also real and useful. Google describes Gemma 4 as an open model family built from the same research and technology as Gemini, designed to run from cloud servers down to laptops and phones. That is valuable. But it is explicitly positioned as a complement to Gemini, not a replacement for it. That is the gap. America has open models. **It does not have an open frontier default.** Gemini could be that default. The objection is obvious: why would Google give away its best model? Because the thing it would gain may be more valuable than the thing it would lose. Google does not need Gemini to be a tollbooth. Google needs Gemini to be the standard. If the weights are open, Google still sells the best hosted version on Vertex AI. It still sells enterprise support, security, monitoring, fine-tuning, deployment, compliance wrappers, managed inference, and integration into Workspace and Cloud. Red Hat made money on Linux. Databricks made money on Spark. Google can make money hosting, securing, accelerating, and operationalizing the model everyone wants to use. The model weights would be the distribution strategy. Cloud would be the monetization layer. The second objection is safety, and yeah, sure, it is serious. But this is coming either way Open frontier weights cannot be easily recalled. Guardrails can be modified or removed. Abuse becomes harder to monitor. Anyone pretending otherwise is not being honest. But that argument has a shelf life, and it is expiring. Open-weight frontier-class models already exist. The question is no longer whether powerful open models will be available. They are. The question is whose models become the global base layer. If Google withholds Gemini, it does not prevent open frontier AI from spreading. It only ensures that the most important open models are built somewhere else. That also matters because access itself is becoming geopolitical. Recent reporting has already shown how governments may pressure access to powerful AI systems, and how both the U.S. and China are thinking about controls around advanced models. A closed API can be throttled, blocked, sanctioned, priced up, or switched off. Open weights cannot be revoked once they are downloaded. That permanence is exactly why enterprises, startups, governments, and research labs care about them. They want models they can inspect, host, air-gap, fine-tune, and keep. If America wants to lead open AI, it has to release something worth building on. That is why Gemini matters. Google has the research history, the infrastructure, the distribution, the enterprise relationships, the cloud, the developer ecosystem, and the brand memory to pull this off. It is the only U.S. company that can open a true frontier model and plausibly turn that act into a platform victory instead of a business-model suicide note. Anthropic cannot do this easily. Its business depends on closed-model trust and premium access. OpenAI has opened smaller models, but its frontier strategy remains closed. Meta could try, but its open-model lead has been challenged and they are now rolling out Spark. Google is different. Google has always been strongest when it turned core technology into infrastructure and let the world build on top. The talent story only makes the argument sharper. Noam Shazeer, one of the original Transformer paper authors and a Gemini co-lead, left Google for OpenAI in 2026 after Google had reportedly paid billions in 2024 to bring him and part of Character.AI’s team back. John Jumper, the Nobel-winning AlphaFold scientist, also left Google DeepMind for Anthropic. A few departures obviously does not kill a lab. Google DeepMind is still loaded with talent. But symbolism matters. When the people associated with your greatest AI achievements keep leaving for the companies you are chasing, the answer is not to keep playing defense. The answer is to change the battlefield. Open sourcing Gemini would do that. It would turn Google from another closed-model vendor into the company that saved the American open-weight ecosystem. It would make Gemini the default base model for developers who do not want to build on Chinese weights. It would give enterprises a domestic model they can self-host. It would give Washington a national competitiveness win without having to design one. It would give Google Cloud a massive inference and deployment wedge. It would make every fine-tune, wrapper, benchmark, agent framework, and deployment guide a piece of unpaid Google distribution. Most importantly, it would make Google feel like Google again. The old Google did not win by hiding every important idea behind a meter. It won by making itself unavoidable. It published the paper. It built the browser. It gave away the mobile OS. It created tools the world adopted because the tools were too useful not to adopt. The AI race does not need another closed API with a better launch video. It needs an American open-weight model strong enough to become the global default. Google can keep Gemini closed and continue fighting OpenAI and Anthropic model by model, benchmark by benchmark, delay by delay. Or it can do the one thing those companies cannot easily match: release the weights and turn Gemini into infrastructure. Google already gave the world the blueprint once. Now it should give the world the model and establish America as the Open Weights leader