Topics
Latest
AI
Amazon
Image Credits:Larysa Amosova via Getty
Apps
Biotech & Health
Climate
Image Credits:Larysa Amosova via Getty
Cloud Computing
mercantilism
Crypto
Enterprise
EVs
Fintech
fund raise
Gadgets
gage
Government & Policy
Hardware
Layoffs
Media & Entertainment
Meta
Microsoft
privateness
Robotics
surety
societal
Space
Startups
TikTok
Transportation
Venture
More from TechCrunch
Events
Startup Battlefield
StrictlyVC
Podcasts
Videos
Partner Content
TechCrunch Brand Studio
Crunchboard
meet Us
In a brief message Tuesday morn on Threads , Meta CEO Mark Zuckerberg said the company ’s “ exposed ” AI model family , Llama , hit 1 billion downloads . That ’s up from 650 million downloads as of early December 2024 — a ~53 % increase over a approximately three - month stop .
Llama , which power Meta ’s AI help , Meta AI , across the tech giant ’s various political program , including Facebook , Instagram , and WhatsApp , is a part of Meta ’s yearslong bid to foster a wide - array AI production ecosystem . The company make the model , as well as the tools required to fine - melodic line and customize them , uncommitted for free under a proprietary license .
Some developer and companies have taken issue with theLlama license terms , which are passably commercially restrictive . Yet Llama has achieved widespread winner since launch in 2023 . Companies including Spotify , AT&T , and DoorDash use Llama models in product today .
That ’s not to suggest that Meta has n’t faced reverse .
Llama is at the centerfield of anAI copyright lawsuitthat accuses Meta of training a routine of exemplar on copyrighted e - playscript without license . Several EU countrieshave impel Meta to hold over — and in some cases cancel altogether — its Llama launch plans over data privacy concerns . And Llama ’s performance has been leapfrogged by model like Formosan AI lab DeepSeek’sR1 .
Meta is say to have scrambled to gear up up “ warfare elbow room ” to apply DeepSeek ’s learnings to Llama ’s own exploitation , and the company recently said it wouldspend as much as $ 80 billionon projects related to AI this year .
Metais design to launchseveral Llama modelling over the next few month , include “ reasoning ” modelling along the lines of OpenAI’so3 - miniand models with natively multimodal capability . Zuckerberg has alsohinted at “ agentic ” features , suggesting that some of these models will be capable to take actions autonomously .
Join us at TechCrunch Sessions: AI
Exhibit at TechCrunch Sessions: AI
“ I call up this very well could be the year when Llama and open source become the most advanced and widely used AI models , ” Zuckerberg allege during Meta ’s Q4 2024 profit call in January . “ [ O]ur goal for [ Llama this year ] is to take . ”
We ’re sure to see more atLlamaCon , Meta ’s first productive AI developer group discussion , which is schedule to take station on April 29 .
