What are the hardware requirements for Opt 30B?

Opt 30B requires approximately 60.1 GB of VRAM and supports a context window of 2K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Opt 30B and how large is it?

Opt 30B is developed by facebook, a model with 30b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

How does Opt 30B perform on standard benchmarks?

Opt 30B has the following published scores: MMLU 26.66. Compare against reference models on this page or on the LLM Explorer leaderboards.

Opt 30B by facebook — VRAM 60.1GB, 2K context

Name: Opt 30B
Author: facebook

Opt 30B is an open-source language model by facebook. Features: 30b LLM, VRAM: 60.1GB, Context: 2K, License: other, LLM Explorer Score: 0.1, Arc: 43.3, HellaSwag: 74.1, MMLU: 26.7, GSM8K: 2.2.

Arxiv:2005.14165 Arxiv:2205.01068 En Jax Opt Pytorch Region:us Sharded Tf

Model Card on HF 🤗: https://huggingface.co/facebook/opt-30b

Opt 30B Parameters and Internals

Model Type

decoder-only, causal language modeling

Use Cases

Primary Use Cases:

text generation, zero- and few-shot learning

Limitations:

bias and safety concerns, quality issues in generation diversity and hallucination

Considerations:

Use of unfiltered internet content can introduce bias.

Additional Notes

OPT models range from 125M to 175B parameters.

Supported Languages

English (predominantly), Non-English (small amount)

Training Details

Data Sources:

BookCorpus, CC-Stories, The Pile, Pushshift.io Reddit, CCNewsV2

Data Volume:

180B tokens

Methodology:

causal language modeling (CLM) objective

Context Length:

2048

Training Time:

~33 days

Hardware Used:

992 80GB A100 GPUs

Model Architecture:

decoder-only transformers

Input Output

Input Format:

tokenized using GPT2 byte-level BPE

Accepted Modalities:

text

Output Format:

text

Performance Tips:

Use half-precision on GPU for large models.

LLM Name	Opt 30B
Repository 🤗	https://huggingface.co/facebook/opt-30b
Model Size	30b
Required VRAM	60.1 GB
Updated	2026-07-18
Maintainer	facebook
Model Type	opt
Model Files	9.8 GB: 1-of-7 9.9 GB: 2-of-7 9.9 GB: 3-of-7 9.9 GB: 4-of-7 9.9 GB: 5-of-7 9.9 GB: 6-of-7 0.8 GB: 7-of-7
Supported Languages	en
Model Architecture	OPTForCausalLM
License	other
Context Length	2048
Model Max Length	2048
Transformers Version	4.21.0.dev0
Beginning of Sentence Token	</s>
End of Sentence Token	</s>
Unk Token	</s>
Vocabulary Size	50272
Torch Data Type	float16
Activation Function	relu
Errors	replace

Rank the Opt 30B capabilities

Have you tried this model? Rate its performance — this feedback helps the ML community find the right model for their needs.

Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation

Best Alternatives to Opt 30B

Best Alternatives	Context / RAM	Downloads	Likes
Galpaca 30B MiniOrca	2K / 59.6 GB	48	1
Galpaca 30B	2K / 60.8 GB	59	55
Opt Iml Max 30B	2K / 60.1 GB	1150	35
OPT 30B Erebus	2K / 36 GB	60	67
...alactica 30B Evol Instruct 70K	2K / 60.1 GB	57	23
Opt Iml 30B	2K / 60.1 GB	36	73
Galactica 30B	2K / 60.8 GB	71	40
Dalio Io 30B Cp1500	2K / 121.7 GB	11	0
Galactica 30B	2K / GB	19	1
...h 1e 5 Lr Handwritten Complete	2K / 121.7 GB	18	0

Note: green Score (e.g. "73.2") means that the model is better than facebook/opt-30b.