M
Llama 3.2 3B InstructOPEN
Llama 3.2 3B is a 3-billion-parameter multilingual large language model, optimized for advanced natural language processing tasks like dialogue generation, reasoning, and summarization.
Prices update automatically — checked daily against provider list prices.
- $0.12 per 1M tokens, assuming 3 input for 1 output the convention used by Artificial Analysis
- cheaper than 75% of the 191 open-weight models tracked here
- 3 of 12 Meta models tracked here, from cheapest
Benchmarks & performance
Independent benchmark scores for Llama 3.2 3B Instruct, measured by Artificial Analysis. Higher is better.
Models at this level cost $0.5 per 1M tokens (median of 11) — this one costs $0.12.
Cheaper and better on this index: DeepSeek V4 Flash 0731 · DeepSeek V4 Flash 0423 · Solar Pro 4 and 17 more
About this model
Llama 3.2 3B Instruct is an open-weight AI model by Meta. You can download and self-host it for free; the prices below are hosted-API list prices, tracked daily, for when you prefer convenience over self-hosting.
Frequently asked questions
What is Llama 3.2 3B Instruct?
Llama 3.2 3B Instruct is an AI language model from Meta. It is open-weight: you can download it and run it on your own hardware, for free. It scores 3.9 on the Artificial Analysis intelligence index.
Is Llama 3.2 3B Instruct free?
The weights are free and open — you can self-host Llama 3.2 3B Instruct and pay nothing per token. If you prefer a hosted API, list prices are $0.05 per million input tokens and $0.33 per million output tokens.
What is Llama 3.2 3B Instruct good at?
Independent benchmarks from Artificial Analysis give it GPQA 25.5%, MMLU-Pro 34.7%, Humanity's Last Exam 5.3%, Long Context Reasoning 2.7%, LiveCodeBench 8.3%, SciCode 5.2%, MATH-500 48.9%, AIME 6.7%, AIME 2025 3.3%, IFBench 26.2%, τ²-Bench 21.1%. It is particularly used for mathematical reasoning.
Can I self-host Llama 3.2 3B Instruct?
Yes. Llama 3.2 3B Instruct has open weights, so you can download it and run it on your own GPU or server with tools like Ollama, vLLM or llama.cpp — with no per-token cost.