No longer supported

This model is legacy (released February 2023) and is no longer actively maintained or recommended by Meta. Consider using their current flagship model instead.

legacylegacyllm Open Source

Llama 1

by Meta· Released February 2023· Cutoff September 2022

Llama 1 is Meta's first open-source large language model, released in February 2023. It was designed to democratize access to LLMs for researchers and developers, offering sizes from 7B to 65B parameters. Llama 1 set a new standard for open-source LLMs, achieving strong performance on benchmarks despite its relatively smaller size compared to proprietary models.

Official Site 🤗 Hugging Face 📄 Research Paper

Input cost

Free (open source)

Output cost

Free (open source)

Context window

2048 tokens

Max output

—

Modalities

text

Parameters

7B, 13B, 33B, 65B

License

Llama 1 License (research-only, non-commercial)

Capabilities

Text GenerationCode GenerationQuestion AnsweringSummarizationTranslationFew-shot Learning

Best For

Research and experimentation with open-source LLMs, especially for tasks requiring strong few-shot performance.

Strengths

Strong performance relative to model size
Open-source and freely available
Multiple sizes for different compute budgets
Good few-shot learning capabilities

Limitations

Limited context window of 2048 tokens
No instruction tuning (base model)
No multimodal capabilities
Outperformed by later Llama models

Use Cases

Academic research on LLMs

Fine-tuning for specific domains

Text generation and completion

Code generation and analysis

Language translation

Summarization of documents

Question answering systems

Back to all models

No longer supported

This model is legacy (released February 2023) and is no longer actively maintained or recommended by Meta. Consider using their current flagship model instead.

legacylegacyllm Open Source

Llama 1

by Meta· Released February 2023· Cutoff September 2022

Official Site 🤗 Hugging Face 📄 Research Paper

Input cost

Free (open source)

Output cost

Free (open source)

Context window

2048 tokens

Max output

—

Modalities

text

Parameters

7B, 13B, 33B, 65B

License

Llama 1 License (research-only, non-commercial)

Capabilities

Text GenerationCode GenerationQuestion AnsweringSummarizationTranslationFew-shot Learning

Best For

Research and experimentation with open-source LLMs, especially for tasks requiring strong few-shot performance.

Strengths

Strong performance relative to model size
Open-source and freely available
Multiple sizes for different compute budgets
Good few-shot learning capabilities

Limitations

Limited context window of 2048 tokens
No instruction tuning (base model)
No multimodal capabilities
Outperformed by later Llama models

Use Cases

Academic research on LLMs

Fine-tuning for specific domains

Text generation and completion

Code generation and analysis

Language translation

Summarization of documents

Question answering systems

Back to all models