What are the hardware requirements for BigCodeLlama 169B?

BigCodeLlama 169B requires approximately 338.2 GB of VRAM and supports a context window of 16K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed BigCodeLlama 169B and how large is it?

BigCodeLlama 169B is developed by nisten, a model with 70b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

Where can I download or evaluate BigCodeLlama 169B?

BigCodeLlama 169B is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

BigCodeLlama 169B by nisten — VRAM 338.2GB, 16K context

Name: BigCodeLlama 169B
Author: nisten

BigCodeLlama 169B is an open-source language model by nisten. Features: 70b LLM, VRAM: 338.2GB, Context: 16K, License: mit, Code Generating, LLM Explorer Score: 0.11.

Base model:codellama/codellama... Base model:finetune:codellama/... Code Codegen Conversational Endpoints compatible Llama Merge Mergekit Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/nisten/BigCodeLlama-169b

BigCodeLlama 169B Parameters and Internals

Model Type

conversational

Additional Notes

This is an experimental CodeLlaMA frankenstein model designed to explore benchmarking results.

Training Details

Methodology:

The model is constructed using a passthrough merge method with bfloat16 data type. It integrates slices from various instances of the CodeLlama model series, specifically focusing on different layer ranges from different variants of CodeLlama.

LLM Name	BigCodeLlama 169B
Repository 🤗	https://huggingface.co/nisten/BigCodeLlama-169b
Base Model(s)	CodeLlama 70B Hf codellama/CodeLlama-70b-hf
Model Size	70b
Required VRAM	338.2 GB
Updated	2026-04-06
Maintainer	nisten
Model Type	llama
Model Files	9.8 GB: 1-of-35 9.8 GB: 2-of-35 9.8 GB: 3-of-35 9.6 GB: 4-of-35 9.8 GB: 5-of-35 9.8 GB: 6-of-35 10.0 GB: 7-of-35 10.0 GB: 8-of-35 9.9 GB: 9-of-35 9.8 GB: 10-of-35 10.0 GB: 11-of-35 9.8 GB: 12-of-35 9.8 GB: 13-of-35 10.0 GB: 14-of-35 9.8 GB: 15-of-35 9.8 GB: 16-of-35 10.0 GB: 17-of-35 9.9 GB: 18-of-35 9.8 GB: 19-of-35 9.8 GB: 20-of-35 10.0 GB: 21-of-35 9.8 GB: 22-of-35 9.9 GB: 23-of-35 9.6 GB: 24-of-35 9.6 GB: 25-of-35 10.0 GB: 26-of-35 9.8 GB: 27-of-35 10.0 GB: 28-of-35 9.7 GB: 29-of-35 9.8 GB: 30-of-35 9.8 GB: 31-of-35 9.9 GB: 32-of-35 9.8 GB: 33-of-35 9.7 GB: 34-of-35 3.8 GB: 35-of-35
Generates Code	Yes
Model Architecture	LlamaForCausalLM
License	mit
Context Length	16384
Model Max Length	16384
Transformers Version	4.37.2
Tokenizer Class	LlamaTokenizer
Vocabulary Size	32016
Torch Data Type	bfloat16

Rank the BigCodeLlama 169B 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 BigCodeLlama 169B

Best Alternatives	Context / RAM	Downloads	Likes
CodeLlama 70B Hf	16K / 77 GB	629	324
Sqlcoder 70B Alpha	16K / 138.7 GB	30	232
CodeLlama 70B Hf	16K / 77 GB	42	29
...enbuddy Codellama 70B V17.1 4K	16K / 138.9 GB	9	1
...enbuddy Codellama 70B V17.1 4K	16K / 138.9 GB	5	1
CodeLlama 70B Instruct Hf	4K / 72.3 GB	729	24
Code Llama 70B Python Instruct	4K / 138.1 GB	5	1
CodeLlama 70B Instruct Hf	4K / 72.3 GB	463	210
CodeLlama 70B Python Hf	4K / 77 GB	186	111
CodeRosa 70B AB1	4K / 61.3 GB	70	1

Note: green Score (e.g. "73.2") means that the model is better than nisten/BigCodeLlama-169b.