Google caps Meta's Gemini use as AI capacity constraints hit even tech giants

Google caps Meta's Gemini use as AI demand strains capacity

Google has put limits on Meta's use of its Gemini AI models after the social media giant sought more computing capacity than the rival tech group could provide, in the latest evidence of the infrastructure constraints facing even the world's largest AI providers. Google told Meta around March that it could not provide all of the Gemini capacity the company wanted to purchase, according to three people familiar with the matter, in a move that has disrupted and delayed some of Meta's internal AI projects. Owing to the restrictions, which remain in place, as well as a broader push to streamline AI costs, Meta has encouraged staff to be more efficient with AI tokens -- the units that measure AI usage, several people said. Several other Google clients have been affected by the restrictions, although to a lesser extent, according to one person familiar with the matter. Meta has been particularly impacted because of its exceptionally high demand for Google's models, the person said. The decision by Google to cap a large customer's access to its models offers a rare glimpse into the infrastructure pressures and bottlenecks building across the AI industry. Despite spending tens of billions of dollars on chips, data centres and power, even the largest tech companies are struggling to secure enough computing power to support surging demand for advanced models and AI services. As a direct result of the demands, particularly from big corporate customers such as Meta, Google has raced to secure additional capacity, according to one person familiar with the matter. Google earlier this month signed a $920mn-a-month deal to lease computing capacity from Elon Musk's SpaceX. Google and Meta declined to comment. At its first-quarter earnings in April, Google chief executive Sundar Pichai said that the company's cloud revenue exceeded $20bn for the first time, while its backlog of signed -- but not yet delivered -- cloud contracts nearly doubled quarter on quarter to more than $460bn. "Obviously, we are compute-constrained in the near term," Pichai said. "And as an example, our Cloud revenue would have been higher if we were able to meet the demand." Demand for AI computing has risen sharply as companies deploy chatbots, coding assistants and AI agents across their businesses. The resulting increase in inference workloads -- tasks required to run models after they have been trained -- has emerged as one of the industry's biggest challenges. AI lab Anthropic, the maker of the popular Claude chatbot, last month struck a deal with SpaceX that is similar to the deal it has with Google. The constraints illustrate the extent to which Meta has relied on rival models such as Gemini, as the social platform spends aggressively to become a leader in AI and improve its own models. Chief executive Mark Zuckerberg has been pouring billions of dollars into tapping talent and securing infrastructure in order to develop what he dubs "personal superintelligence". Unlike Google, Meta does not have a cloud business and is racing to build out its fleet of data centres for its own training and inference needs. As part of the push, Meta has committed to investing $600bn in the US by 2028. Gemini has been used internally at Meta as part of a push to automate some of its safety processes, such as rooting out scams and taking down harmful content, as well as for its customer services and advertising help chatbots. It is also used internally for some workflows and coding, alongside other models such as Anthropic's Claude. Meta initially chose to use Gemini because it performed better than the social media company's own Llama open-source models, according to people familiar with the matter. More recently, Meta has begun to shift to prioritise its new Muse Spark model, several people said, which is viewed as more competitive with Gemini and reduces the company's dependence on external models for some applications.

Google limits Meta's use of its Gemini AI models, FT reports

June 28 (Reuters) - Google has put limits on Meta's (META.O), opens new tab use of its Gemini AI models after the social media company sought more computing capacity than the rival tech group could provide, the Financial Times reported on Sunday. Google, owned by Alphabet (GOOGL.O), opens new tab, told Meta around March it could not meet the full Gemini capacity the company had sought to purchase, the newspaper said, ⁠adding that the shortfall disrupted and delayed some of Meta's internal AI projects. Several other Google clients have also been affected, though to a lesser extent, according to the report. Meta has been particularly impacted due to its exceptionally high demand for Google's models, the FT said. Reuters could not immediately verify the report, which cited people familiar with the matter. Google and Meta did not ⁠immediately respond to requests for comment outside business hours. Due to the restrictions, Meta has encouraged staff to be more efficient with AI tokens, the units that measure AI usage, the FT report said. Even as ⁠companies continue to spend billions on chips and data centres, they are still struggling to secure enough computing power to support the growing demand ⁠for AI services. Revenue at Google Cloud grew to $20 billion in the first quarter ended March, but CEO Sundar Pichai said computing ⁠power constraints prevented even higher growth and contributed to the cloud unit's backlog nearly doubling quarter on quarter. Reporting by Abu Sultan in Bengaluru; Editing by William Mallard and Sonali Paul Our Standards: The Thomson Reuters Trust Principles., opens new tab

Google reportedly capped Meta's use of Gemini AI for coding and chatbots - Engadget

Even tech giants with their own LLMs are having trouble finding enough computing power. Google was forced to cap Meta's use of its Gemini AI model after Mark Zuckerberg's company exceeded its computing capacity, sources familiar with the matter told The Financial Times. The incident reveals that even tech giants with their own LLMs are having trouble finding enough computing power for themselves, let alone their customers. Meta does not operate its own cloud business and is trying to rapidly expand its own data center build out, having pledged $600 billion in cloud computing investments over the next two years. Google reportedly warned the social media company about its capacity limits in March, in turn forcing Meta to request that employees use tokens more efficiently, the sources said. Gemini AI is being used by Meta for customer service, advertiser chatbots and coding, alongside processes like harmful content takedowns and scam detection. Meta initially chose Gemini for those things because it outperformed its own Llama open-source models, the sources said. The company also employs other models like Anthropic's Claude for similar purposes. Despite the billions spent so far on data centers, big companies are struggling to get enough capacity for their usage needs. Google itself recently agreed to pay SpaceX $920 million a month to use xAI's data centers, due to the extra computing power required for Gemini Enterprise. AI power users are benefiting from the boom, but providers like OpenAI aren't profiting yet, since revenue earned from AI so far is a small percentage of the costs, according to analysts. Recently, token prices have surged, forcing some companies to back off on AI usage -- including, it appears, the AI companies themselves.

Meta has been secretly relying on Google's AI for customer service, ad tools, and content moderation - then got cut off

Serving tech enthusiasts for over 25 years. TechSpot means tech analysis and advice you can trust. WTF?! Few companies seem less likely to run out of AI capacity than Google and Meta, but even the industry's biggest names can hit a token wall. The search giant has reportedly limited the Facebook owner's use of Gemini after demand for AI capacity grew beyond what it could supply. According to the Financial Times, Google warned Meta around March that it could not provide all the capacity the company wanted, disrupting and delaying internal AI projects. The restrictions are still said to be in place. Meta has reportedly told employees to be more careful with AI tokens, the units used to measure model input, output, and usage. That's quite the change of tone for a company that has spent the past year pushing - and in some cases forcing - staff to use AI as much as possible. Meta has spent billions building its own Llama family of open models, while Mark Zuckerberg has been pitching AI as the company's next defining platform, one Meta will hope does not go the same way as its metaverse bet. But people familiar with the arrangement told the FT that Meta had been using Google's Gemini models for customer service, advertiser chatbots, coding, harmful content takedowns, and scam detection. Gemini was reportedly chosen because it performed better than Meta's own models. Anthropic's Claude is also said to be in the mix. The shortage didn't hit only Meta; other Google customers were also affected, though less severely. Meta appears to have been the outlier because of the amount of Gemini capacity it wanted to buy. The reliance is not entirely surprising. Meta doesn't operate a cloud business of its own, unlike Google, Microsoft, or Amazon, leaving it to balance its internal AI systems with outside capacity from the same companies it competes against. Google has been spending heavily on data centers and AI hardware, but demand is still arriving faster than capacity can be built. Google Cloud revenue passed $20 billion in Alphabet's most recent quarter, while backlog nearly doubled to more than $460 billion. The company also said its first-party models were processing more than 16 billion tokens per minute through direct API use, up 60% from the previous quarter. Meta is trying to solve the same problem, but from the other direction. It's been expanding data centers and working with Broadcom on custom MTIA accelerators as it looks to rely less on rivals.

Meta Reportedly Got Too Addicted to Google AI Tokens and Had to Be Cut Off

Meta was reportedly minding its own business this past March, just trying to gorge itself on Gemini tokens, and all of a sudden Google said it was cut off. This is according to an anonymously sourced story in the Financial Times. In March, it emerged that Meta was one of the largest companies taking part in the Tokenmaxxing trend -- literally evaluating employees by how many AI tokens they were using at work. This moment coincided with another fad: the one for token-hungry agentic AI platforms like OpenClaw, which were being used by anxious software engineers to achieve ostensibly unprecedented new levels of workplace efficiency. Citing "three people familiar with the matter," the Financial Times now says Google informed Meta that it wasn't able to keep up with its AI use, and imposed limits on the company's use of Gemini models. The move has, FT says: "...disrupted and delayed some of Meta's internal AI projects. Owing to the restrictions, which remain in place, as well as a broader push to streamline AI costs, Meta has encouraged staff to be more efficient with AI tokens -- the units that measure AI usage, several people said." In other words, Meta replaced tokenmaxxing with judicious token-counting. Sad! The burden on Google's resources, meanwhile, could have helped along Google's decision in early June to rent compute from SpaceX, the parent company of xAI for $920 million per month. The FT says other large companies also strained Google's AI capacity and were subject to caps, but it sounds like those problems weren't as serious. According to the FT's sources, Meta was exceptional, even among the other AI high-rollers. Meta and Google declined the FT's requests for comment.

Google is rationing Gemini access to Meta because it cannot provide enough compute

Google capped Meta's Gemini access due to compute constraints. Meta told staff to use AI tokens more efficiently and is shifting to its own Muse Spark model. Google has placed limits on Meta's use of its Gemini AI models because it cannot provide as much computing capacity as the social media company wanted, the Financial Times reported on Sunday. The restrictions have affected several Google clients, with Meta hit particularly hard. The move has had a knock-on effect on Meta's internal projects. The company has told staff to make more efficient use of AI tokens, according to three people familiar with the matter cited by the FT. Both Google and Meta declined to comment. Meta had initially relied on Gemini, which proved better than its own Llama open-source models, to automate safety processes like removing harmful content and wiping out scams. It has increasingly been shifting workloads to Muse Spark, a new internal model, as it looks to reduce dependence on external AI providers. Google itself is so compute-constrained that it agreed to pay SpaceX $920 million a month for access to 110,000 Nvidia GPUs, calling it "bridge capacity" to meet surging demand for Gemini Enterprise. The situation illustrates how the AI compute shortage is reshaping relationships between the industry's largest companies. Google, which owns one of the world's largest pools of AI infrastructure and is spending over $180 billion on capex this year, still cannot serve all of its customers' demand. That it is rationing access to a company as large as Meta, while simultaneously renting GPUs from a rocket company, is the clearest signal yet that AI infrastructure buildouts have not kept pace with consumption. For Meta, the dependence on a competitor's AI models was always an uncomfortable arrangement. The company cut 8,000 jobs in May and redirected billions toward AI infrastructure, with capex guidance of $115 to $135 billion for 2026. It has reassigned 7,000 workers to AI-focused roles and launched Muse Spark under its Superintelligence Labs division. The Gemini restrictions accelerate a transition Meta was already pursuing, from relying on external frontier models to building internal alternatives capable of handling critical workloads like content moderation at scale. The broader pattern is consistent across the industry. Demand for AI compute is growing faster than even the most aggressive infrastructure spending can supply. Google is buying capacity from SpaceX. Anthropic is renting an entire data centre from SpaceX. Meta is being told to use fewer tokens by its own cloud provider. The AI boom's most tangible bottleneck is not algorithms or talent. It is the physical infrastructure required to run them.

Google reportedly limits Meta Gemini access over compute shortage

Google has limited Meta's access to its Gemini AI models due to compute constraints, severely impacting the social media company, the Financial Times reported. As a result, Meta has announced a shift toward its internal Muse Spark model to reduce dependence on external providers. The restrictions forced Meta to instruct staff to use AI tokens more efficiently, according to three sources cited by the Financial Times. Both Google and Meta declined to comment on the situation. Meta had been relying on Gemini to automate safety processes, including content moderation and scam removal. The shift towards Muse Spark aims to alleviate reliance on Google's AI resources amidst ongoing compute shortages. Google's own computing limitations have led the company to pay SpaceX $920 million a month for access to 110,000 Nvidia GPUs, referred to as "bridge capacity" for its Gemini Enterprise. This underscores how the current AI compute shortage is affecting the relationships among major companies in the sector. Despite Google's investments in AI infrastructure totaling over $180 billion this year, it has not been able to meet all customer demand. The company is rationing access to customers like Meta while simultaneously securing GPU capacity from SpaceX. Meta's situation reflects its ongoing transition from dependency on its competitor's AI models to developing internal alternatives. In May, Meta laid off 8,000 employees and redirected significant resources toward its own AI infrastructure, projecting capital expenditures between $115 billion to $135 billion for 2026. The company has reassigned 7,000 workers to AI-focused roles and recently launched the Muse Spark model under its Superintelligence Labs. This transition aligns with a broader industry trend where demand for AI compute continues to outstrip the capacity provided by major players. Companies like Anthropic are also looking for solutions, such as renting data centers from SpaceX, to meet their AI operational needs. The overall pattern indicates that the physical infrastructure required to support AI algorithms and talent remains the bottleneck in the AI boom, surpassing all prior expectations for infrastructure spending.

Google Limits Meta's Access To Gemini AI Models Amid Rising Demand: Report - Alphabet (NASDAQ:GOOGL)

Google Informed Meta In March About Limitations Google informed Meta in March about the limitations, which have led to disruptions and delays in Meta's internal AI initiatives, according to a report by the Financial Times. The restrictions continue to remain in place, prompting Meta to encourage its staff to optimize AI token usage, according to sources familiar with the situation. Other Google Clients Facing Similar Restrictions Other Google clients have also faced similar restrictions, though to a lesser degree. Meta's significant demand for Google's models has made it particularly vulnerable to these constraints, one source noted, according to the report. Google's move to cap access highlights the infrastructure challenges facing the AI industry. Despite substantial investments in technology, major companies like Google struggle to meet the growing demand for AI services. Google has recently secured additional capacity, including a $920 million monthly deal with SpaceX for computing resources. Meta has been using Gemini for automating safety processes and enhancing customer services. However, the company is now prioritizing its Muse Spark model to reduce reliance on external models. Evolving Google-Meta Relationship The relationship between Google and Meta has been evolving over the years, marked by significant collaborations and agreements. In August last year, Meta struck a $10 billion cloud pact with Google to bolster its AI capabilities, despite their competitive rivalry. Earlier this month, Meta launched a new AI tool within Facebook Search, powered by its Muse Spark model, which is expected to potentially generate revenue of $10 billion annually. Photo courtesy: Shutterstock Disclaimer: This content was partially produced with the help of AI tools and was reviewed and published by Benzinga editors. Market News and Data brought to you by Benzinga APIs To add Benzinga News as your preferred source on Google, click here.

Why Google moved Gemini to token-based limits

Google has tightened how Gemini's free and paid tiers work. Since May 17, 2026, Gemini Apps have run with compute-based usage limits based on prompt complexity, the model chosen, and the length of the chat, refreshing every five hours up to a weekly cap. The change is the consumer-facing edge of a compute shortage that has squeezed Google's biggest enterprise customers for months, and it is reshaping how much free AI ordinary users get. Who got hit first? Google capped Meta's use of its Gemini models in March after Meta sought more capacity than Google could supply, the Financial Times reported. The restriction, still in place, disrupted some of Meta's internal AI projects and pushed the company to tell staff to use AI tokens, the units that measure AI usage, more efficiently. The constraints hit several other Google clients too, though less hard than Meta. Capping one of its largest customers signals how serious the shortage already was. Has Google admitted the problem? At its first-quarter earnings in April, CEO Sundar Pichai said Google was "compute-constrained in the near term," and that cloud revenue would have been higher had it been able to meet demand, per the FT. Cloud revenue crossed $20 billion for the first time, while signed-but-undelivered cloud contracts nearly doubled to more than $460 billion. That backlog is the tell: customers are signing up for compute faster than Google can deliver it. How are providers coping? Google signed a $920 million-a-month deal to lease computing capacity from Elon Musk's SpaceX, and Anthropic, which makes Claude, struck a similar arrangement with SpaceX. These leasing deals show the shortage has become a structural feature of the market, which is why the limits are unlikely to loosen soon. What is driving the cost? The highest cost now comes from inference, the work of running models after training, the FT noted. Every prompt a user sends consumes compute, so the more people use AI for everyday tasks, the heavier that running cost becomes. Does Google link the limits to capacity? Google's Gemini Apps help page ties the limits to capacity directly: * If capacity changes, Google may cut limits for free users before paying subscribers. * During periods of high demand, Google may withhold compute-heavy features such as Deep Research from free users. * Usage limits may change without notice, including due to capacity constraints, and Google may tighten them when activity spikes. By Google's own account, capacity sets the limits. Where do consumers fit in? Token-based limits let a provider ration finite compute across its user base. People who never touch an API use Gemini Apps as mass-market tools to summarise, brainstorm and generate images. As that everyday usage scales, it puts the same constrained infrastructure under greater strain. Metering by compute lets Google charge heavier tasks more and steer demand towards lighter models, matching each user's cost to their actual consumption. What does each tier get? What a user gets now scales with what they pay: * Context window: 32,000 tokens on the free tier, 128,000 on AI Plus, one million on AI Pro and AI Ultra. * Usage limits: standard on the free tier, 2x on Plus, 4x on Pro, and up to 20x on the top Ultra plan. * Gated features: video generation, scheduled actions and the Daily Brief sit behind paid plans. What does this mean for India? Google built its Indian user base on free access, and that access is now governed by compute-based usage limits. A wave of free giveaways onboarded Indian users to AI, with three deals landing close together in late 2025: * Google-Jio: 18 months of free Google AI Pro, a plan worth about Rs 35,100, for eligible subscribers. * OpenAI: ChatGPT Go offered free to Indian users. * Perplexity-Airtel: a free Perplexity Pro subscription for the telco's customers. MediaNama founder Nikhil Pahwa read the wave as a "scale first, monetise later" play, the same free-access land grab Jio used to capture the telecom market. Compute-based limits are what the monetise phase looks like in practice. Can India supply its own compute? India's flagship IndiaAI Mission set aside roughly Rs 4,563 crore for compute capacity, against more than $52 billion in private AI infrastructure commitments from Amazon and Microsoft alone, MediaNama reported. At a MediaNama roundtable, Ajay Kumar of Triumvir Law argued that India should incentivise the private sector to build domestic compute capacity so that "one day we're not at the mercy of foreign data centers" for running public infrastructure. When the global providers ration capacity, the markets that depend on them feel it first. Why it matters. The Gemini change tells consumers something the enterprise story already showed: AI economics are tightening for everyone. More generous free access is giving way to tighter metered limits, because the underlying resource is genuinely scarce and providers are paying heavily to secure it. For users, more of what AI can do will sit behind paid tiers. The deeper question is who controls the compute on which all of this runs, who can afford access to it, and on what terms.

Google Limits Meta's Gemini AI Access Amid Rising Compute Demand

Google has reportedly limited Meta's access to Gemini AI models due to compute constraints, highlighting growing pressure on AI infrastructure as tech giants compete for processing power to expand generative AI capabilities. Google has reportedly restricted Meta's access to its Gemini AI models after the social media giant requested more computing capacity than Google could provide. The move highlights the mounting infrastructure challenges even the world's biggest AI companies are facing as demand for generative AI continues to rise. According to a Financial Times report, Google informed Meta around March that it could not fulfill the company's full request for Gemini capacity, disrupting some of Meta's internal AI projects. Reuters has not independently verified the report.

Google limits Meta's Gemini AI access as compute demand outpaces supply By Investing.com

Investing.com -- Google has restricted Meta Platforms' access to its Gemini artificial intelligence models after the Facebook parent sought more computing capacity than Google could provide, the Financial Times reported on Sunday, citing people familiar with the matter. According to the report, Google informed Meta around March that it could not meet all of the company's requested Gemini capacity. The restrictions remain in place and have delayed some of Meta's internal AI projects. Meta has also encouraged employees to use AI resources more efficiently as part of a broader effort to reduce computing costs. Other Google customers have reportedly been affected by similar capacity limits, though the impact has been greatest on Meta because of its unusually high demand. The move highlights growing infrastructure constraints across the AI industry, where surging demand for computing power continues to outpace available capacity despite heavy investment in chips, data centers, and power infrastructure. Google has been expanding its own computing resources to address rising demand. Earlier this month, the company agreed to lease computing capacity from SpaceX in a deal reportedly worth about $920 million per month. Speaking during the company's first-quarter earnings call in April, Chief Executive Sundar Pichai said Google Cloud revenue exceeded $20 billion for the first time, while the backlog of signed but undelivered cloud contracts nearly doubled from the previous quarter to more than $460 billion. Pichai also acknowledged that computing capacity remains constrained in the near term, adding that cloud revenue would have been higher if Google had been able to meet customer demand. Meta has been investing heavily in artificial intelligence infrastructure as Chief Executive Mark Zuckerberg seeks to strengthen the company's AI capabilities. The company has committed to investing up to $600 billion in the United States through 2028 to expand data center capacity. Internally, Meta has used Google's Gemini models for coding, customer service, advertising tools, and content moderation. The report said the company has recently begun shifting some workloads to its own Muse Spark model, reducing its reliance on third-party AI models for certain applications.

Google reportedly limits Meta's Gemini access: compute crunch delays work

Since about March, Google has reportedly been putting limits on Meta Platforms' access to Gemini for a pretty basic reason: Google couldn't spare the amount of compute Meta wanted. Those limits reportedly started around March, affected multiple Google Cloud customers, and landed hardest on Meta Platforms because Meta's demand is so large. The result was slower internal model work and Meta teams being told to cut token use, even though Gemini reportedly outperformed some of Meta's Llama models on moderation, harmful-content removal, and scam detection. So the bottleneck is right there in the open: GPUs, memory, power. That also helps explain why Meta Platforms is shifting more workloads to its in-house Muse Spark under Meta Superintelligence Labs, rerouting billions toward infrastructure, guiding to $115 billion to $135 billion in 2026 capex, and moving 7,000 workers into model roles. Anyone following AI platforms should pay attention to this. Even Google still looks tight on capacity, and that's with Alphabet CEO Sundar Pichai acknowledging growth pressure inside Google Cloud, Google's Q1 backlog nearly doubling, more than $180 billion in capex, and a reported $920 million-a-month SpaceX deal for roughly 110,000 Nvidia GPUs. Gemini is available through Google Cloud, while Meta Platforms is moving work onto Muse Spark. Consumers may not notice much right away. Smaller firms might, especially as the platforms with the most compute decide who ships on time, who gets better faster, and who gets to sell the infrastructure in the first place.

Google reportedly limits Meta's access to Gemini AI models, here is why

Google informed Meta around March that it could not provide all the Gemini AI capacity the company wanted to buy. Google has reportedly placed limits on how much of its Gemini AI models Meta can use. This decision is said to come after the social media giant requested more computing power than Google could provide. This highlights a growing problem in the AI industry that there is not enough computing power to meet demand. According to a report by the Financial Times, Google informed Meta around March that it could not provide all the Gemini AI capacity the company wanted to buy. As a result, some of Meta's internal AI projects have been delayed. Meta has reportedly asked its employees to use AI tokens more carefully. The report also says that Meta is not the only company facing this issue. Some other Google customers have also experienced limits on AI computing resources, but their impact has reportedly been smaller. Meta has been affected more because of its unusually high demand for Google's AI models. Also read: Apple iPhone 18 Pro series and iPhone Ultra may launch on this date: Here is what we know This shows that even some of the world's biggest technology companies are struggling to keep up with the growth of AI. Companies continue to invest billions of dollars in new data centres and advanced AI chips, but building this infrastructure takes time. Google has already acknowledged that computing capacity is becoming a challenge. During the first-quarter earnings announcement, Google CEO Sundar Pichai said that demand for Google Cloud services was so strong that limited computing resources prevented the business from growing even faster. Also read: OpenAI unveils GPT 5.6 family of AI models, but you can't use them yet: Here is why Just last month, OpenAI introduced a new offering called Guaranteed Capacity for customers who need to buy access to AI computing power for their products, AI agents and workflows. The company said that the offering "enables customers to guarantee long-term access to OpenAI compute." Customers can choose between one-year, two-year or three-year commitments. OpenAI CEO Sam Altman explained the reason behind the offering, saying that the "customers are increasingly asking us for certainty on capacity. as models get better, we expect that the world will be capacity-constrained for some time."