Meta unveils Llama 3.3 70B: High performance with reduced resource usage
Meta has launched the language model Llama 3.3 70B. The model is reportedly designed to match the performance of the significantly larger 405B version, but at a lower computational cost.
What happened?
Meta has released the AI model Llama 3.3 70B, an open-weights language model with 70 billion parameters. According to Meta, the model performs at a level comparable to the previous and significantly larger Llama 3.1 405B on several benchmarks, while operating with substantially lower computational and overhead costs. The model has been made available for download via platforms such as Hugging Face and Meta’s official developer portal.
Key facts
| Modellnamn | Llama 3.3 70B |
|---|---|
| Antal parametrar | 70 miljarder (70B) |
| Tillgänglighet | Hugging Face, Meta Llama webbplats |
| Planerad datacentersatsning | 10 miljarder USD |
Why it matters
The efficiency improvements in Llama 3.3 mean that advanced language model capabilities are now accessible to a broader range of actors at a lower infrastructure cost. Meta claims that in internal tests, Llama 3.3 70B competes with or outperforms proprietary models such as Google Gemini 1.5 Pro, OpenAI GPT-4o, and Amazon Nova Pro on specific benchmarks.
Who is affected?
Llama 3.3 70B is primarily aimed at developers, researchers, and enterprises that wish to run advanced language models on their own hardware or in private cloud environments without the resource requirements of the 400B-class models. Marketers and analysts may also utilise the model for text generation, data analysis, and automation.
Impact on the EU
The Llama 3.3 70B model is available via open platforms such as Hugging Face, making its weights accessible globally. Usage within the EU is subject to the EU AI Act and applicable data protection legislation such as the GDPR, which sets requirements for how personal data is handled during fine-tuning and commercial operation.
What else you should know
Meta also stated that it plans significant investments in its computational infrastructure, including a new 10-billion-dollar data centre, to facilitate the training of the upcoming model generation, Llama 4.
Quick answers about this story
Vad har hänt?
När hände det?
Varför spelar det roll?
Gäller detta i EU och Sverige?
The link opens in a new window and leads to the publisher's own site.
Källan har spårats automatiskt från utgivaren via Aheadlines signalkedja.
AI-verktyg i artikeln
Topics
Get similar news straight to your inbox
The reader's room
Send in a question or an addition. The newsroom reads everything before it's published and replies when relevant. No AI-generated text – just people.
Sign in to submit a comment or question.
Read the article through your role
- Decide whether this affects strategy over 6–12 months or is just noise.
- Discuss with leadership: do we own the right question or does ownership need to move?
- Ask: what risk are we taking by NOT acting on this this quarter?
Generated angle — not editorial analysis of "Meta unveils Llama 3.3 70B: High performance with reduced re"