Google introduces new class of cheap AI models as cost concerns intensify
SAN FRANCISCO, Feb 5 - Alphabet's (GOOGL.O), opens new tab Google on Wednesday announced updates to its Gemini family of large language models, including a new product line with competitive pricing to low-cost artificial intelligence models like that of Chinese rival DeepSeek.
The tech giant offers several versions of Gemini that vary in price and performance. It already offered a lightweight variant known as "Flash" but its new "Flash-Lite" model is even cheaper.
On Wednesday, Google released Gemini 2.0 Flash to the general public after previewing it to developers in December. It also launched Flash-Lite and released a new version of its flagship "Pro" model into test phases.
Google created Flash-Lite after receiving positive feedback about the 1.5 version of Flash, Koray Kavukcuoglu, chief technology officer of Google's DeepMind AI lab, said in a press release. Gemini 2.0 Flash is costlier than its predecessor.
The cost to develop AI models and, in turn, the cost to use them have come under investor scrutiny in recent weeks after DeepSeek revealed it spent less than $6 million on the final training run of a model. Developers at leading U.S. AI firms said the total cost was likely magnitudes larger.
Still, DeepSeek's rise drew questions on the earnings calls of Alphabet and rivals Microsoft (MSFT.O), opens new tab and Meta (META.O), opens new tab. All have so far signaled intent to continue massive capital expenditures in the field.
Alphabet shares slumped Tuesday in part due to investor pessimism around a planned capex hike that was 29% higher than Wall Street expected.
Certain inputs on Gemini Flash-Lite cost $0.019 per 1 million tokens, a term for the units of data processed by an AI model. That compares to $0.075 on the cost-efficient version of OpenAI's flagship model and $0.014 on DeepSeek's cheap model, though DeepSeek states on its website that the pricing will increase fivefold on Feb. 8.
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Reuters
7 hours ago
- Reuters
CoreWeave to offer compute capacity in Google's new cloud deal with OpenAI, sources say
June 11 (Reuters) - CoreWeave (CRWV.O), opens new tab has emerged as a winner in Google's newly signed partnership with OpenAI, sources familiar with the matter told Reuters, in the latest example of the voracious appetite for computing resources in the artificial-intelligence industry and the formation of new alliances to meet them. The so-called neocloud company, which sells cloud computing services built on Nvidia's (NVDA.O), opens new tab graphics processing units, is slated to provide computing capacity to Google's cloud unit, and Alphabet's (GOOGL.O), opens new tab Google will then sell that computing capacity to OpenAI to meet the growing demand for services like ChatGPT, the sources said. Google will also provide some of its own computing resources to OpenAI, added the sources, who requested anonymity to discuss private matters. The details of the arrangement, first reported by Reuters on Tuesday, highlight the evolving dynamics between hyperscalers like (AMZN.O), opens new tab, Microsoft (MSFT.O), opens new tab and Google and so-called neocloud companies like Coreweave. Hyperscalers are large cloud service providers that offer massive-scale data centers and cloud infrastructure. The insatiable hunger for computing resources has generated major investment commitments and turned rivals into partners. Backed by OpenAI and Nvidia, Coreweave signed up Google as a customer in the first quarter. CoreWeave, Google and OpenAI declined to comment. CoreWeave, a specialized cloud provider that went public in March, has already been a major supplier of OpenAI's infrastructure. It has signed a five-year contract worth $11.9 billion with OpenAI to provide dedicated computing capacity for OpenAI's model training and inference. OpenAI also took a $350 million equity stake in CoreWeave in March. This partnership was further expanded last month through an additional agreement worth up to $4 billion, extending through April 2029, underscoring OpenAI's escalating demand for high-performance computing resources. Industry insiders say adding Google Cloud as a new customer could help CoreWeave diversify its revenue sources, and having a credible partner with deep pockets like Google enables the startup to secure more favorable financing terms to support ambitious data center buildouts across the country. This could also boost Google's cloud unit, which generated $43 billion in sales last year, allowing it to capitalize on the growth of OpenAI, which is also one of its largest competitors in areas like search and chatbots. It positions Google as a neutral provider of computing resources in competition with peers such as Amazon and Microsoft. CoreWeave's deal with Google coincides with Microsoft's re-evaluation of its data center strategy, including withdrawing from certain data center leases. Microsoft, once Coreweave's largest customer, accounting for about 62% of its 2024 revenue, is also renegotiating with OpenAI to revise the terms of their multibillion-dollar investment, including the future equity stake it will hold in OpenAI. CoreWeave, backed by Nvidia, has established itself as a fast-rising provider of GPU-based cloud infrastructure in the AI wave. While its public debut in March was met with a lukewarm response due to concerns over its highly leveraged capital structure and shifting GPU demand, the company's stock has surged since its IPO price of $40 per share, gaining over 270% and reaching a record high of $166.63 in June.
Auto Blog
7 hours ago
- Auto Blog
www.autoblog.com
Check any cables and reboot any routers, modems, or other network devices you may be using. Allow Chrome to access the network in your firewall or antivirus settings. If it is already listed as a program allowed to access the network, try removing it from the list and adding it again.
Reuters
9 hours ago
- Reuters
Google names insider Kavukcuoglu to lead AI-powered product development
June 11 (Reuters) - Alphabet's (GOOGL.O), opens new tab Google on Wednesday appointed an insider to lead its future AI-powered product development, as AI enters a new phase of mainstream adoption, according to a memo from CEO Sundar Pichai to company leaders seen by Reuters. Koray Kavukcuoglu, chief technology officer of Google's DeepMind AI lab, is taking on the newly created title of chief AI architect. He will also become a new senior vice president, reporting directly to Pichai. Kavukcuoglu, who will move to California from London, will also continue serving as Google DeepMind's CTO, reporting to its CEO Demis Hassabis, Pichai said in the memo. "In this expanded role, Koray will accelerate how we bring our world-leading models into our products, with the goal of more seamless integration, faster iteration, and greater efficiency," Pichai said. Alphabet faces market pressure to demonstrate financial returns on its AI-related capital expenditures, which are expected to hit $75 billion this year, while maintaining its bottom line against the threat of competing AI offerings, as well as antitrust enforcement. Google said it would put artificial intelligence into the hands of more web surfers and teased a $249.99-a-month subscription for its AI power users at its annual I/O conference in May, along with a flurry of demos that included new smart glasses. Pichai said in Wednesday's memo that Alphabet was "entering a new phase of the AI platform shift." He told reporters last month that the rise of generative AI was not at the full expense of online search. Google has since signed, opens new tab a deal to supply computing capacity to its biggest AI rival OpenAI via its cloud service, Reuters exclusively reported on Tuesday, marking a surprising collaboration between two prominent competitors in the AI sector.
