Sapling Logo

Llama 3.1 vs. Llama 3.2

LLM Comparison


Llama 3.1

Llama 3.1

Overview

Llama 3.1 added a 405B flagship model that was the first open-source model competitive with frontier proprietary models, alongside updated 8B and 70B variants with longer context and improved tool use.


Llama 3.1 was a landmark release for open-source AI. The 405B model achieved parity with GPT-4o and Claude 3.5 Sonnet on many benchmarks, marking the first time an open model reached the proprietary frontier. All models support a 128K token context window, multilingual generation in eight languages, and built-in tool-calling capabilities. The permissive license allows commercial use for most organisations, making Llama 3.1 a popular foundation for fine-tuning and self-hosted deployments.


Initial release: 2024-07-23

Llama 3.2

Llama 3.2

Overview

Llama 3.2 introduced multimodal vision models (11B and 90B) alongside lightweight text-only models (1B and 3B) designed for on-device and edge deployment.


Llama 3.2 expanded Meta's open-source model family in two directions. The vision models (11B and 90B) are the first Llama models to support image understanding, enabling tasks such as chart analysis, document parsing, and visual question answering. The lightweight text models (1B and 3B) are optimised for on-device inference, with quantised versions available for mobile and edge hardware. All models benefit from the architectural improvements introduced in Llama 3.1.


Initial release: 2024-09-25

Looking for an LLM API/SDK that works out of the box? No prompts or ad hoc guardrails.

Sapling API
More Comparisons

Llama 3.1

Llama 3.2
Products & Features
Instruct Models
Coding Capability
Customization
Finetuning
Open Source
License Custom (Commercial OK) Custom (Commercial OK)
Model Sizes 8B, 70B, 405B 1B, 3B, 11B, 90B

Other Llama 3.1 Comparisons

Llama 3.1 vs. Alpaca Llama 3.1 vs. Cerebras-GPT Llama 3.1 vs. DeepSeek Llama 3.1 vs. Dolly Llama 3.1 vs. Falcon Llama 3.1 vs. FastChat Llama 3.1 vs. FLAN-T5 Llama 3.1 vs. FLAN-UL2 Llama 3.1 vs. Gemma Llama 3.1 vs. Gemma 2 Llama 3.1 vs. GPT-J Llama 3.1 vs. GPT4All Llama 3.1 vs. GPTNeo Llama 3.1 vs. Grok Llama 3.1 vs. Guanaco Llama 3.1 vs. Koala Llama 3.1 vs. LLaMA Llama 3.1 vs. Llama 2 Llama 3.1 vs. Llama 3 Llama 3.1 vs. Llama 3.2 Llama 3.1 vs. Llama 3.3 Llama 3.1 vs. Mistral Llama 3.1 vs. MPT Llama 3.1 vs. OpenAssistant Llama 3.1 vs. OpenLLaMA Llama 3.1 vs. OPT Llama 3.1 vs. Orca Llama 3.1 vs. Phi Llama 3.1 vs. Pythia Llama 3.1 vs. Qwen Llama 3.1 vs. RedPajama-INCITE Llama 3.1 vs. StableLM Llama 3.1 vs. StableVicuna Llama 3.1 vs. Vicuna Llama 3.1 vs. WizardLM

Other Llama 3.2 Comparisons

Llama 3.2 vs. Alpaca Llama 3.2 vs. Cerebras-GPT Llama 3.2 vs. DeepSeek Llama 3.2 vs. Dolly Llama 3.2 vs. Falcon Llama 3.2 vs. FastChat Llama 3.2 vs. FLAN-T5 Llama 3.2 vs. FLAN-UL2 Llama 3.2 vs. Gemma Llama 3.2 vs. Gemma 2 Llama 3.2 vs. GPT-J Llama 3.2 vs. GPT4All Llama 3.2 vs. GPTNeo Llama 3.2 vs. Grok Llama 3.2 vs. Guanaco Llama 3.2 vs. Koala Llama 3.2 vs. LLaMA Llama 3.2 vs. Llama 2 Llama 3.2 vs. Llama 3 Llama 3.2 vs. Llama 3.1 Llama 3.2 vs. Llama 3.3 Llama 3.2 vs. Mistral Llama 3.2 vs. MPT Llama 3.2 vs. OpenAssistant Llama 3.2 vs. OpenLLaMA Llama 3.2 vs. OPT Llama 3.2 vs. Orca Llama 3.2 vs. Phi Llama 3.2 vs. Pythia Llama 3.2 vs. Qwen Llama 3.2 vs. RedPajama-INCITE Llama 3.2 vs. StableLM Llama 3.2 vs. StableVicuna Llama 3.2 vs. Vicuna Llama 3.2 vs. WizardLM

Undecided?

Get Advice