Qwen2.5-7B-Instruct by Qwen: Open-Weight Model

Runs On

What it takes to serve Qwen2.5-7B-Instruct (7.6B parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them.

Precision	Weights	Memory needed	Cheapest setup	Per hour	Also fits
16-bit	15.2 GB	18.3 GB	1x MI300X (192 GB) Vultr	$1.85	1x H100 $1.99 · 1x MI325X $2.00
8-bit	7.6 GB	9.1 GB	1x MI300X (192 GB) Vultr	$1.85	1x H100 $1.99 · 1x MI325X $2.00
4-bit	3.8 GB	4.6 GB	1x MI300X (192 GB) Vultr	$1.85	1x H100 $1.99 · 1x MI325X $2.00

Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the SAVRN Index, read Sep 18, 2026.

SAVRN's Notes on Qwen2.5-7B-Instruct

Structured work is what Qwen built this 7.6B instruct model for: reading tables, writing JSON and producing text past 8K tokens, in their words. At 16-bit the run needs 18.3 GB, and the cheapest listed setup, one MI300X with 192 GB at $1.85 an hour on-demand, fits ten copies by simple arithmetic. At 4-bit the need falls to 4.6 GB, small enough to share a card with other services.

Apache 2.0 lets you ship it in a product, fine-tuned or not, with notices kept and significant changes stated. It derives from Qwen/Qwen2.5-7B, so the base is there if you would rather tune it yourself. The page lists 32,768 tokens of context and cites arXiv:2309.00071, the YaRN context-extension paper, so confirm which window your serving stack honors before promising long documents. No host quotes it by the token on the SAVRN Index today; the card hour is your cost basis.

SAVRN Research, 2026-09-18

Model Card

By Qwen, published under apache-2.0, revision a09a35458c70.

Introduction

Qwen2.5 is the latest series of Qwen large language models. For Qwen2.5, we release a number of base language models and instruction-tuned language models ranging from 0.5 to 72 billion parameters. Qwen2.5 brings the following improvements upon Qwen2:

Significantly more knowledge and has greatly improved capabilities in coding and mathematics, thanks to our specialized expert models in these domains.
Significant improvements in instruction following, generating long texts (over 8K tokens), understanding structured data (e.g, tables), and generating structured outputs especially JSON. More resilient to the diversity of system prompts, enhancing role-play implementation and condition-setting for chatbots.
Long-context Support up to 128K tokens and can generate up to 8K tokens.
Multilingual support for over 29 languages, including Chinese, English, French, Spanish, Portuguese, German, Italian, Russian, Japanese, Korean, Vietnamese, Thai, Arabic, and more.

Read the full model card (753 words)

Configuration

Architecture: Qwen2ForCausalLM
Context length (tokens): 32,768
Layers: 28
Hidden size: 3,584
Feed-forward size: 18,944
Attention heads: 28
Key/value heads: 4
Vocabulary size: 152,064
Sliding window (tokens): 131,072
RoPE base: 1e+06
Stored precision: bfloat16
Model type: qwen2

Identity and Version

Repository: Qwen/Qwen2.5-7B-Instruct
Publisher: Qwen
Task: Text generation
Modality: Text
Library: transformers
Parameters: 7.6B parameters
Languages: en
Revision: a09a35458c702b33eeacc393d103063234e8bc28
First published: 2024-09-16
Last updated: 2025-01-12

Files and Weights

14 files, 15.2 GB in total. The weights are 4 files totalling 15.2 GB in safetensors.

Weights4 files · 15.2 GB

Configuration3 files · 28.7 KB

Tokenizer4 files · 11.5 MB

Documentation2 files · 17.6 KB

Repository1 file · 1.5 KB

Every file

File	Type	Size	SHA-256
model-00001-of-00004.safetensors	Weights	3.9 GB	a1333e629385
model-00002-of-00004.safetensors	Weights	3.9 GB	f5d25a2772cb
model-00003-of-00004.safetensors	Weights	3.9 GB	8efdec4c1bc1
model-00004-of-00004.safetensors	Weights	3.6 GB	1a72d403cdf0
config.json	Configuration	663 B	—
generation_config.json	Configuration	243 B	—
model.safetensors.index.json	Configuration	27.8 KB	—
LICENSE	Documentation	11.3 KB	—
README.md	Documentation	6.2 KB	—
.gitattributes	Repository	1.5 KB	—
merges.txt	Tokenizer	1.7 MB	—
tokenizer.json	Tokenizer	7.0 MB	—
tokenizer_config.json	Tokenizer	7.3 KB	—
vocab.json	Tokenizer	2.8 MB	—

License and Download

License: apache-2.0
Access: Open weights, no gate
Download size: 15.2 GB

Download from Qwen

Released by Qwen through ModelScope. Read the license.

Built From

Derived from Qwen/Qwen2.5-7B
Described by arXiv:2309.00071
Described by arXiv:2407.10671

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.

Benchmark	Conditions	Result	Reported by	Revision	Date
LEXam-Benchmark/LEXam	Task mcq_4_choicesMetric mcq_4_choicesComparison conditions not established	29.28	LEXam Leaderboard Reported by a third party	Evaluated revision not stated	2026-06-02
LEXam-Benchmark/LEXam	Task open_questionMetric open_questionComparison conditions not established	16.67	LEXam Leaderboard Reported by a third party	Evaluated revision not stated	2026-06-02
thamilvendhan/signalbench	Task access_denyMetric access_denySetup family=access_deny; n=12Comparison conditions not established	0.1667	thamilvendhan Reported by a third party	Evaluated revision not stated	2026-07-08
thamilvendhan/signalbench	Task bot_policyMetric bot_policySetup family=bot_policy; n=12Comparison conditions not established	0.3333	thamilvendhan Reported by a third party	Evaluated revision not stated	2026-07-08
thamilvendhan/signalbench	Task injectionMetric injectionSetup family=injection; n=12Comparison conditions not established	0.5	thamilvendhan Reported by a third party	Evaluated revision not stated	2026-07-08
thamilvendhan/signalbench	Task memory_labelMetric memory_labelSetup family=memory_label; n=12Comparison conditions not established	0.5	thamilvendhan Reported by a third party	Evaluated revision not stated	2026-07-08
thamilvendhan/signalbench	Task srcMetric srcSetup SRC overall; deterministic action-based grader, no LLM judge; seed 0, n=75Comparison conditions not established	0.3833	signalbench raw per-item responses Reported by a third party	Evaluated revision not stated	2026-07-08
thamilvendhan/signalbench	Task timeMetric timeSetup family=time; n=12Comparison conditions not established	0.4167	thamilvendhan Reported by a third party	Evaluated revision not stated	2026-07-08

Memory Requirements

Precision	Weights in memory
As published	15.2 GB
16-bit	15.2 GB
8-bit	7.6 GB
4-bit	3.8 GB

Weights only, from the published parameter count; the key-value cache and runtime add to this.

Built on This Model

Quantized fromQwen2.5-7B-Instruct-AWQ
Derived fromQwen2.5-7B-Instruct-AWQ

Compare Qwen2.5-7B-Instruct

Questions About Qwen2.5-7B-Instruct

How much GPU memory does Qwen2.5-7B-Instruct need?

About 18.3 GB at 16-bit and 4.6 GB at 4-bit: the weights (7.6B parameters) plus a working margin. A long context needs more.

What is the cheapest GPU to run Qwen2.5-7B-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 Qwen2.5-7B-Instruct commercially?

Yes. Qwen2.5-7B-Instruct is released under Apache License 2.0. The Apache License 2.0 is a permissive open-source license. It permits commercial use, modification and redistribution. It requires keeping the license and copyright notices and any NOTICE file, stating significant changes, and it includes an express patent grant from contributors.

What is Qwen2.5-7B-Instruct's context length?

32,768 tokens, from the maximum position embeddings in its published configuration.

Similar Models

Model · Text generation

Qwen2.5-7B-Instruct-AWQ

Qwen

Qwen2.5 is the latest series of Qwen large language models. For Qwen2.5, we release a number of base language models and instruction-tuned language models ranging from 0.5 to 72 billion parameters. Qwen2.5 brings the following improvements upon Qwen2: - Significantly more knowledge and has greatly improved capabilities in coding and mathematics, thanks to our specialized expert models in these domains. - Significant improvements in instruction following, generating long texts (over 8K tokens), understanding structured data (e.g, tables), and generating structured outputs especially JSON. More resilient to the diversity of system prompts, enhancing role-play implementation and…

Open weights apache-2.0 7.6B parameters 32,768 tokens transformers

View model

Model · Text generation

Qwen2.5-Coder-7B-Instruct

Qwen

Qwen2.5-Coder is the latest series of Code-Specific Qwen large language models (formerly known as CodeQwen). As of now, Qwen2.5-Coder has covered six mainstream model sizes, 0.5, 1.5, 3, 7, 14, 32 billion parameters, to meet the needs of different developers. Qwen2.5-Coder brings the following improvements upon CodeQwen1.5: - Significantly improvements in code generation, code reasoning and code fixing. Base on the strong Qwen2.5, we scale up the training tokens into 5.5 trillion including source code, text-code grounding, Synthetic data, etc. Qwen2.5-Coder-32B has become the current state-of-the-art open-source codeLLM, with its coding abilities matching those of GPT-4o. - A more…

Open weights apache-2.0 7.6B parameters 32,768 tokens transformers

View model

Model · Text generation

Qwen2.5-Coder-7B-Instruct-AWQ

Qwen

Qwen2.5-Coder is the latest series of Code-Specific Qwen large language models (formerly known as CodeQwen). As of now, Qwen2.5-Coder has covered six mainstream model sizes, 0.5, 1.5, 3, 7, 14, 32 billion parameters, to meet the needs of different developers. Qwen2.5-Coder brings the following improvements upon CodeQwen1.5: - Significantly improvements in code generation, code reasoning and code fixing. Base on the strong Qwen2.5, we scale up the training tokens into 5.5 trillion including source code, text-code grounding, Synthetic data, etc. Qwen2.5-Coder-32B has become the current state-of-the-art open-source codeLLM, with its coding abilities matching those of GPT-4o. - A more…

Open weights apache-2.0 7.6B parameters 32,768 tokens transformers

View model

Model · Text generation

Qwen2.5-Coder-7B

Qwen

Qwen2.5-Coder is the latest series of Code-Specific Qwen large language models (formerly known as CodeQwen). As of now, Qwen2.5-Coder has covered six mainstream model sizes, 0.5, 1.5, 3, 7, 14, 32 billion parameters, to meet the needs of different developers. Qwen2.5-Coder brings the following improvements upon CodeQwen1.5: - Significantly improvements in code generation, code reasoning and code fixing. Base on the strong Qwen2.5, we scale up the training tokens into 5.5 trillion including source code, text-code grounding, Synthetic data, etc. Qwen2.5-Coder-32B has become the current state-of-the-art open-source codeLLM, with its coding abilities matching those of GPT-4o. - A more…

Open weights apache-2.0 7.6B parameters 32,768 tokens transformers

View model

Model · Text generation

Qwen2.5-7B

Qwen

Qwen2.5 is the latest series of Qwen large language models. For Qwen2.5, we release a number of base language models and instruction-tuned language models ranging from 0.5 to 72 billion parameters. Qwen2.5 brings the following improvements upon Qwen2: - Significantly more knowledge and has greatly improved capabilities in coding and mathematics, thanks to our specialized expert models in these domains. - Significant improvements in instruction following, generating long texts (over 8K tokens), understanding structured data (e.g, tables), and generating structured outputs especially JSON. More resilient to the diversity of system prompts, enhancing role-play implementation and…

Open weights apache-2.0 7.6B parameters 131,072 tokens transformers

View model

Model · Text generation

JiRackUltra_7b

Center Business Solutions inc

A fast and efficient 7B model optimized for CPU inference. The model was refactored with BitNet features and an updated tokenizer that includes new Routing, Media, Vision, Sound,Tool call, and Robotics tags. Built on a DeepSeek R1 -7B architecture with native ternary (BitNet-style) support and ready-to-run GGUF quantizations. - JiRack is a cloud-ready model that helps save money on cloud infrastructure. It can be used as an expert model in RAG deployments, with the ONNX JiRack Java server as an alternative. - Benefits high quality CPU inference TQ2 on Llama.cpp and Ollama via QAT - Robotcs, Routing, Coding, Multimedia, Advanced tool calling via CMSManhattan/JiRackPrecisionTokenizer…

Open weights mit 7.6B parameters 131,072 tokens

View model