SAVRN's Notes on Meta-Llama-3-8B-Instruct
Nineteen point three gigabytes is the number to plan around at 16-bit, 16.1 of that being weights. At 8-bit the need falls to 9.6 gigabytes; at 4-bit it is 4.8, small enough that the 192 gigabyte MI300X the Index lists as cheapest, at $1.85 an hour, is far more card than one copy needs. That headroom is the point: room to serve many sessions at once.
The Llama 3 Community License permits commercial use with conditions: attribution as Meta specifies, compliance with Meta's Acceptable Use Policy, and a separate license request for any licensee with more than 700 million monthly active users on the release date. Access is gated, so the files arrive only after the publisher approves you. Check the context length, which our record does not carry, and the age of the weights, released April 17, 2024 and last updated June 18, 2025.
SAVRN Research, 2026-09-18
Model Card
Meta developed and released the Meta Llama 3 family of large language models (LLMs), a collection of pretrained and instruction tuned generative text models in 8 and 70B sizes. The Llama 3 instruction tuned models are optimized for dialogue use cases and outperform many of the available open source chat models on common industry benchmarks. Further, in developing these models, we took great care to optimize helpfulness and safety. Model developers Meta Variations Llama 3 comes in two sizes — 8B and 70B parameters — in pre-trained and instruction tuned variants. Input Models input text only. Output Models generate text and code only. Model Architecture Llama 3 is an auto-regressive language…
Excerpt from the card by Meta Llama, licensed llama3.
Identity and Version
- Repository
- meta-llama/Meta-Llama-3-8B-Instruct
- Publisher
- Meta Llama
- Task
- Text generation
- Modality
- Text
- Library
- transformers
- Parameters
- 8B parameters
- Languages
- en
- Revision
- 8afb486c1db24fe5011ec46dfbe5b5dccdb575c2
- First published
- 2024-04-17
- Last updated
- 2025-06-18
Files and Weights
17 files, 32.1 GB in total. The weights are 5 files totalling 32.1 GB in pth, safetensors.
Weights5 files · 32.1 GB
Configuration5 files · 25.1 KB
Tokenizer3 files · 11.3 MB
Documentation3 files · 51.4 KB
Repository1 file · 1.5 KB
Every file
License and Download
- License
- llama3
- Access
- Access requested at publisher
- Download size
- 32.1 GB
Evaluations
Each result is shown as reported, with the conditions its reporter stated. None is a SAVRN measurement. A comparison lines two results up only when their configuration, unit and setup are all stated and identical.
Memory Requirements
Weights only, from the published parameter count; the key-value cache and runtime add to this.
Compare Meta-Llama-3-8B-Instruct
Questions About Meta-Llama-3-8B-Instruct
How much GPU memory does Meta-Llama-3-8B-Instruct need?
About 19.3 GB at 16-bit and 4.8 GB at 4-bit: the weights (8B parameters) plus a working margin. A long context needs more.
What is the cheapest GPU to run Meta-Llama-3-8B-Instruct on?
At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists.
Can I use Meta-Llama-3-8B-Instruct commercially?
Yes, with conditions. Meta-Llama-3-8B-Instruct is released under Meta Llama 3 Community License. The Llama 3 Community License permits commercial use, except that a licensee whose products had more than 700 million monthly active users on the release date must request a license from Meta. It requires attribution as the license specifies and compliance with Meta's Acceptable Use Policy.
Similar Models
The Meta Llama 3.1 collection of multilingual large language models (LLMs) is a collection of pretrained and instruction tuned generative models in 8B, 70B and 405B sizes (text in/text out). The Llama 3.1 instruction tuned text only models (8B, 70B, 405B) are optimized for multilingual dialogue use cases and outperform many of the available open source and closed chat models on common industry benchmarks. Model Architecture: Llama 3.1 is an auto-regressive language model that uses an optimized transformer architecture. The tuned versions use supervised fine-tuning (SFT) and reinforcement learning with human feedback (RLHF) to align with human preferences for helpfulness and safety.…
Access requested at publisher
llama3.1
8B parameters
transformers
The Model mlx-community/Llama-3.1-8B-Instruct-4bit was converted to MLX format from meta-llama/Llama-3.1-8B-Instruct using mlx-lm version 0.21.4.
Open weights
llama3.1
8B parameters
131,072 tokens
mlx
This model is a fine-tuned version of unsloth/Llama-3.1-8B-unsloth-bnb-4bit. It has been trained using TRL. This model was trained with SFT.
Open weights
8B parameters
131,072 tokens
transformers
This is the 8B (high-capacity flagship) member of the Med-LLaMA3 family introduced in the paper “Med-LLaMA3: Advancing Medical Question-Answering Through Parameter-Efficient Fine-Tuning of Large Language Models” (Applied Sciences, 2026). The family adapts the LLaMA-3 architecture to the medical domain by training only a small fraction of the base model’s parameters (4.01% for this 8B variant), achieving strong medical question-answering performance while keeping the memory footprint low — enabling development and inference on low-cost, consumer-grade hardware. The 8B variant is the high-capacity model for complex clinical reasoning. It attains a mean accuracy of 75.71% across the eight MMLU…
Open weights
llama3.1
8B parameters
131,072 tokens
transformers
This model is a fine-tuned version of unsloth/Llama-3.1-8B-unsloth-bnb-4bit. It has been trained using TRL. This model was trained with SFT.
Open weights
8B parameters
131,072 tokens
transformers
Lebanese-Llama-3.1-8B is a high-performance LLM fine-tuned specifically for the Lebanese dialect (Ammiya). It bridges the gap between Modern Standard Arabic (MSA) and the multi-modal nature of Lebanese communication, seamlessly blending Arabic script, French/English influences, and Arabizi (Romanized Arabic with numbers). This model was trained and validated on the NVIDIA DGX Spark, the world’s first personal AI supercomputer powered by the Grace Blackwell (GB10) architecture. Try the model instantly in your browser without any setup: To get the most authentic "Ammiya" experience, use this system prompt. It activates the model's specialized cultural knowledge and linguistic patterns. Speak…
Open weights
mit
8B parameters
131,072 tokens
transformers
