What are the hardware requirements for AquilaChat2 34B?

AquilaChat2 34B requires approximately 67.2 GB of VRAM and supports a context window of 4K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed AquilaChat2 34B and how large is it?

AquilaChat2 34B is developed by BAAI, a model with 34b 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 AquilaChat2 34B?

AquilaChat2 34B is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

AquilaChat2 34B by BAAI — VRAM 67.2GB, 4K context

Name: AquilaChat2 34B
Author: BAAI

AquilaChat2 34B is an open-source language model by BAAI. Features: 34b LLM, VRAM: 67.2GB, Context: 4K, License: other, LLM Explorer Score: 0.09.

Arxiv:2408.07410 Aquila Custom code Pytorch Region:us Sharded

Model Card on HF 🤗: https://huggingface.co/BAAI/AquilaChat2-34B

AquilaChat2 34B Parameters and Internals

Model Type

chat model, language model

Use Cases

Areas:

research, commercial applications

Limitations:

data leakage problem identified with GSM8K test data included in training dataset

Supported Languages

en (high), zh (high)

Training Details

Data Sources:

WTM22 (en-zh), CLUEWSC, Winograd, HellaSwag, OpenBookQA, PIQA, ARC-e, BUSTSM, BoolQ, TruthfulQA, RAFT, ChID, EPRSTMT, TNEWS, OCNLI, SEM-Chinese, MMLU, C-Eval, CMMLU, CSL, HumanEval

Data Volume:

2 trillion tokens

Context Length:

16000

Safety Evaluation

Findings:

data leakage issue identified with GSM8K

Risk Categories:

misinformation, bias

Input Output

Input Format:

text input

Accepted Modalities:

text

Output Format:

text output

Release Notes

Version:

1.2

Date:

2023-10-25

Notes:

AquilaChat2-34B model is based on the previous version and exceeds or matches GPT3.5 in various evaluation dimensions.

LLM Name	AquilaChat2 34B
Repository 🤗	https://huggingface.co/BAAI/AquilaChat2-34B
Model Size	34b
Required VRAM	67.2 GB
Updated	2026-08-19
Maintainer	BAAI
Model Type	aquila
Model Files	10.0 GB: 1-of-7 9.8 GB: 2-of-7 9.7 GB: 3-of-7 9.7 GB: 4-of-7 9.7 GB: 5-of-7 9.7 GB: 6-of-7 8.6 GB: 7-of-7
Model Architecture	AquilaForCausalLM
License	other
Context Length	4096
Model Max Length	4096
Transformers Version	4.31.0
Tokenizer Class	GPT2Tokenizer
Vocabulary Size	100008
Torch Data Type	bfloat16

Rank the AquilaChat2 34B 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

Quantized Models of the AquilaChat2 34B

Model	Likes	Downloads	VRAM
AquilaChat2 34B GGUF	3	187	14 GB
AquilaChat2 34B GPTQ	2	16	18 GB
AquilaChat2 34B AWQ	1	6	19 GB

Best Alternatives to AquilaChat2 34B

Best Alternatives	Context / RAM	Downloads	Likes
AquilaChat2 34B 16K	16K / 67.2 GB	176	26
Aquila2 34B	8K / 136.4 GB	306	19
H2ogpt 16K Aquilachat2 34B	4K / 67.2 GB	9	4
AquilaChat2 34B 16K GPTQ	4K / 18.7 GB	7	5
AquilaChat2 34B GPTQ	4K / 18.7 GB	16	2
AquilaChat2 34B 16K AWQ	4K / 19.3 GB	10	3
AquilaChat2 34B AWQ	4K / 19.3 GB	6	1

Note: green Score (e.g. "73.2") means that the model is better than BAAI/AquilaChat2-34B.