What are the hardware requirements for BigCodeLlama 92B?

BigCodeLlama 92B requires approximately 184 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 BigCodeLlama 92B and how large is it?

BigCodeLlama 92B 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.

How does BigCodeLlama 92B perform on standard benchmarks?

BigCodeLlama 92B has the following published scores: MMLU 55.4. Compare against reference models on this page or on the LLM Explorer leaderboards.

BigCodeLlama 92B by nisten — VRAM 184GB, 2K context

Name: BigCodeLlama 92B
Author: nisten

BigCodeLlama 92B is an open-source language model by nisten. Features: 70b LLM, VRAM: 184GB, Context: 2K, License: mit, Instruction-Based, Code Generating, LLM Explorer Score: 0.13, Arc: 54.8, HellaSwag: 77.8, MMLU: 55.4, GSM8K: 45.

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

Model Card on HF 🤗: https://huggingface.co/nisten/BigCodeLlama-92b

BigCodeLlama 92B Parameters and Internals

Model Type

conversational

Additional Notes

This is an experimental 92B parameter model constructed by merging with `../CodeLlama-70b-Python-hf` and `../CodeLlama-70b-Instruct-hf` models using a specific YAML configuration. The model's type is conversational.

LLM Name	BigCodeLlama 92B
Repository 🤗	https://huggingface.co/nisten/BigCodeLlama-92b
Base Model(s)	CodeLlama 70B Instruct Hf codellama/CodeLlama-70b-Instruct-hf
Model Size	70b
Required VRAM	184 GB
Updated	2026-07-06
Maintainer	nisten
Model Type	llama
Instruction-Based	Yes
Model Files	9.8 GB: 1-of-19 9.8 GB: 2-of-19 9.6 GB: 3-of-19 9.8 GB: 4-of-19 9.9 GB: 5-of-19 9.7 GB: 6-of-19 10.0 GB: 7-of-19 9.9 GB: 8-of-19 9.6 GB: 9-of-19 10.0 GB: 10-of-19 9.8 GB: 11-of-19 9.9 GB: 12-of-19 9.8 GB: 13-of-19 9.8 GB: 14-of-19 10.0 GB: 15-of-19 9.9 GB: 16-of-19 9.6 GB: 17-of-19 9.7 GB: 18-of-19 7.4 GB: 19-of-19
Generates Code	Yes
Model Architecture	LlamaForCausalLM
License	mit
Context Length	2048
Model Max Length	2048
Transformers Version	4.37.2
Tokenizer Class	LlamaTokenizer
Vocabulary Size	32016
Torch Data Type	bfloat16

Rank the BigCodeLlama 92B 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 92B

Best Alternatives	Context / RAM	Downloads	Likes
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 Instruct Neuron	4K / GB	5	1
...Llama 70B Instruct Hf 4bit MLX	4K / 39.1 GB	417	25
...70B Instruct Nf4 Fp16 Upscaled	4K / 138.7 GB	10	1
...70B Instruct Hf 5.0bpw H6 EXL2	2K / 43.6 GB	5	6
...70B Instruct Hf 2.4bpw H6 EXL2	2K / 21.3 GB	6	1
...0B Instruct Hf 2.65bpw H6 EXL2	2K / 23.4 GB	5	3
...70B Instruct Hf 4.0bpw H6 EXL2	2K / 35.1 GB	6	1

Note: green Score (e.g. "73.2") means that the model is better than nisten/BigCodeLlama-92b.