Llama
Meta's family of open-weight large language models
What is Llama?
Llama is Meta's family of open-weight large language and multimodal models that organisations can download, fine-tune and run on their own infrastructure or through cloud providers. Models are available in multiple sizes for tasks such as chat, reasoning, code and image understanding. The Llama Stack and API offerings provide tooling for building applications and agents. It is used by developers and enterprises that want control over model deployment and data.
Key facts
- Starting price
- Free
- Pricing model
- Free
- Active users
- —
- Customers
- —
- Vendor employees
- 74,067
- Launched
- 2023
- Deployment
- Cloud (SaaS), On-premise, Hybrid
- Platforms
- Linux, Web
Strengths & considerations
Strengths
- Open-weight models can be self-hosted and fine-tuned
- Large ecosystem of tools, hosting providers and community
- Range of model sizes for different hardware budgets
- Free plan available
- Can run on-premise
Considerations
- Community license has usage conditions to review
- Self-hosting requires GPU infrastructure and ML skills
- No vendor-managed support by default
Editorial summary based on public information; not a user review.
Industries & use cases
Industries served
Technology, media & telecom
Banking & capital markets
Healthcare
Government & public sector
Cross-industry
Typical use cases
- Self-hosted chatbot over sensitive internal data
- Fine-tuned domain models for document processing
- On-device or edge generative AI inference
Typical Deloitte services
- Open-model generative AI architecture and deployment
- Model fine-tuning and evaluation programs
- AI governance for self-hosted models
Editorial profile based on public information.
Features
✓AI features· Automation
✓Self-hosted· Deployment
✓API access· Integrations
✓Multi-language· Platform
User reviews
No reviews yetNo reviews yet. Be the first to write one!