AlibabaQwen 2.5Qwen2ForCausalLM

Qwen2.5 Coder 7B Instruct — Hardware Requirements & GPU Compatibility

ChatCodeFunctions

Qwen2.5 Coder 7B Instruct is a 7.6-billion parameter code-specialized instruction-tuned model from Alibaba Cloud. It is trained on a large corpus of source code and natural language, fine-tuned for programming assistance tasks such as code generation, completion, debugging, and code explanation. The model supports a 128K token context window and runs efficiently on consumer GPUs with 8GB or more of VRAM. It provides a good balance between coding capability and hardware requirements for developers looking to run a local coding assistant. Released under the Apache 2.0 license.

7.6B params · dense33K contextApache 2.0Sep 20241.3M downloads865 likes2.1M quant downloads

Based on Qwen2.5 Coder 7B

Your hardware

Verdicts below are for the reference RTX 4090. Set yours to personalise the page.

Detection reads your browser's GPU name with no data sent anywhere.

Runs great on a 24 GB GPU

Runs from 8 GB (Q4_K_M · 32K)Best on 24 GB (Q4_K_M · ~131 tok/s · 32K)Macs from 8 GB (Q2_K · ~20 tok/s)

Apache 2.0

Reference GPU · RTX 4090 24 GB

Estimates, ±20% · how we estimate

Run Qwen2.5 Coder 7B Instruct

One recommended way. Every other method, OS and quant is in the step-by-step guide.Each way opens a step-by-step guide for your OS and quant.

Ways to runOther ways

each opens step-by-step

Run in cloud

Your own private copy · Q4_K_M · 24 GB GPU

The easiest way to try this model
Running in a few minutes, easy setup
Chat right here or connect any app

$5+ bonus for new users

Run Qwen2.5 Coder 7B Instruct

5.8 GB VRAM at 16K context · fits the RTX 4090 (18.2 GB free)

Works on macOS, Windows and Linux.

Install Ollama (Linux)
$curl -fsSL https://ollama.com/install.sh | sh
Install Ollama (macOS / Windows)
Download and run the app; it installs the ollama command.
Download Ollama ↗
Run Q4_K_M
$ollama run qwen2.5-coder:7b
Official Ollama library build. For the exact Q4_K_M file from Qwen/Qwen2.5-Coder-7B-Instruct-GGUF: ollama run hf.co/Qwen/Qwen2.5-Coder-7B-Instruct-GGUF:Q4_K_M
Raise the context window
Type it inside the chat. For the API server instead, start it with OLLAMA_CONTEXT_LENGTH=16384 ollama serve.
$/set parameter num_ctx 16384
A larger context needs more memory than the weights alone.
Use the API
Ollama serves its API on http://localhost:11434.

Works on macOS, Windows and Linux.

Install LM Studio
Desktop app for macOS (Apple silicon), Windows and Linux.
lmstudio.ai ↗
Download the model
In the app press Ctrl/Cmd + Shift + M and paste https://huggingface.co/Qwen/Qwen2.5-Coder-7B-Instruct-GGUF, then pick Q4_K_M. Or use the CLI (run LM Studio once first):
$lms get https://huggingface.co/Qwen/Qwen2.5-Coder-7B-Instruct-GGUF@Q4_K_M
Open on Hugging Face ↗
Load and chat
Open the model loader, select the model and set Context Length to 16384. Leave GPU offload at maximum if it fits.
Start the local server (optional)
$lms server start

Works on macOS, Windows and Linux.

Install llama.cpp (macOS)
MacPorts, Nix and conda-forge also work.
$brew install llama.cpp
Install options ↗
Install llama.cpp (Windows)
$winget install llama.cpp
Install llama.cpp (Linux)
Download the prebuilt llama-*-bin-ubuntu-* archive for your GPU backend (CPU, Vulkan) from the releases page, unpack it and run the binaries from that folder. Homebrew (brew install llama.cpp) and conda-forge also work on Linux.
llama.cpp releases ↗
Start the server
Downloads the model on first run. The chat UI and an OpenAI-compatible API are at http://localhost:8080.
$llama-server -hf Qwen/Qwen2.5-Coder-7B-Instruct-GGUF:Q4_K_M -c 16384 -ngl 99 --port 8080

Works on macOS, Windows and Linux.

Option A: Docker Model Runner
Docker Desktop: Settings > AI > Enable Docker Model Runner. Docker Engine on Linux: install the docker-model-plugin package.
$docker model run hf.co/Qwen/Qwen2.5-Coder-7B-Instruct-GGUF:Q4_K_M
Model Runner docs ↗
The quant tag after the repo is how Hugging Face cards write it; if your Docker version rejects it, drop the tag and pick the quant by repo.
Option B: llama.cpp server image (NVIDIA)
Needs the NVIDIA Container Toolkit. Models are cached in the llama-cache volume. Open http://localhost:8080.
$docker run --gpus all -p 8080:8080 -v llama-cache:/root/.cache ghcr.io/ggml-org/llama.cpp:server-cuda -hf Qwen/Qwen2.5-Coder-7B-Instruct-GGUF:Q4_K_M -c 16384 -ngl 99 --host 0.0.0.0 --port 8080
Image tags (cuda13, vulkan, rocm) ↗
Cache path inside the image is not in the docs; remove the -v flag if the volume complains.

Linux only.

Install vLLM
Linux only; on Windows run it inside WSL2. NVIDIA GPU shown; AMD ROCm, Intel and CPU builds have their own install pages.
$uv venv --python 3.12 --seed
$source .venv/bin/activate
$uv pip install vllm --torch-backend=auto
vLLM installation ↗
Serve the model
OpenAI-compatible API at http://localhost:8000/v1.
$vllm serve Qwen/Qwen2.5-Coder-7B-Instruct --max-model-len 16384
Runs the original weights, not the Q4_K_M GGUF; needs ~17 GB of GPU memory.

macOS only.

MLX supports only macOS, so follow these steps for that setup.

Install mlx-lm
$pip install mlx-lm
Chat in the terminal
$mlx_lm.chat --model lmstudio-community/Qwen2.5-Coder-7B-Instruct-MLX-4bit
Downloads the model on first run. The MLX repo has its own quantisation; it may not match the quant selected above.
Or start a local server
Listens on http://localhost:8080.
$mlx_lm.server --model lmstudio-community/Qwen2.5-Coder-7B-Instruct-MLX-4bit

Works on macOS, Windows and Linux.

Install Jan
jan.ai ↗
Find the model
Open the Hub tab, paste this model ID in Search, and choose Q4_K_M if offered.
$Qwen/Qwen2.5-Coder-7B-Instruct-GGUF
Use the local API (optional)
Settings > Local API Server > Start Server. It listens on http://127.0.0.1:1337.

Rent GPUs hourly to run Q4_K_M right there. Pricing comes from partner marketplaces and varies.

Available GPU rentals for Q4_K_M, prioritized by value
GPU	Speed	$/1M tokens	Rent
RTX 3080 Ti (12 GB)CheapestCheapest6 GB free	~119 tok/s	$0.29	Deploy on Vast.aifrom $0.12/hr (opens in new tab)	12 GB · 6 GB free · ~119 tok/s · $0.29 / 1M tokens
RTX 3060 12GB6 GB free	~47 tok/s	$0.35	Deploy on Vast.aifrom $0.060/hr (opens in new tab)	6 GB free · ~47 tok/s · $0.35 / 1M tokens
RTX 3060 Ti (8 GB)2 GB free	~58 tok/s	$0.33	Deploy on Vast.aifrom $0.069/hr (opens in new tab)	8 GB · 2 GB free · ~58 tok/s · $0.33 / 1M tokens
RTX 4060 (8 GB)2 GB free	~35 tok/s	$0.67	Deploy on Vast.aifrom $0.086/hr (opens in new tab)	8 GB · 2 GB free · ~35 tok/s · $0.67 / 1M tokens
RTX A4000 (16 GB)10 GB free	~58 tok/s	$0.42	Deploy on Vast.aifrom $0.088/hr (opens in new tab)	16 GB · 10 GB free · ~58 tok/s · $0.42 / 1M tokens
RTX A5000 (24 GB)18 GB free	~100 tok/s	$0.44	Deploy on RunPodfrom $0.16/hr (opens in new tab)	24 GB · 18 GB free · ~100 tok/s · $0.44 / 1M tokens
RTX 4000 Ada (20 GB)14 GB free	~47 tok/s	$1.18	Deploy on RunPodfrom $0.20/hr (opens in new tab)	20 GB · 14 GB free · ~47 tok/s · $1.18 / 1M tokens
RTX 4090 (24 GB)18 GB free	~131 tok/s	$0.72	Deploy on RunPodfrom $0.34/hr (opens in new tab)	24 GB · 18 GB free · ~131 tok/s · $0.72 / 1M tokens

Compare GPUs →

Your own private copy of Qwen2.5 Coder 7B Instruct · Q4_K_M · OpenAI-compatible API

Open the llama.cpp listing on RunPod
Open on RunPodServerless (opens in new tab)
New to RunPod? Sign up and add credit first.
2
Fill in the form
Press Deploy and fill in:
Model
Qwen/Qwen2.5-Coder-7B-Instruct-GGUF
Quantization
q4_k_m
Context Size
16384
After pasting the model press Esc — don’t pick a suggestion. Press Next.
Create the endpoint
Keep the default 24 GB and 48 GB, then Create Endpoint. Right away: Manage → Edit endpoint → Max workers 1 → Save (changing settings later restarts the worker).
The first start downloads the model, usually 5–15 min. After that, replies after a pause come in seconds.
Create an API key
RunPod → Settings → API Keys → Create. Pick “Restricted” with access to this endpoint only.

Use it

Copy the endpoint URL from its page on RunPod (any of its links works) and paste it with the key. Both stay in this browser.

Endpoint URL

API key

Chats and keys stay only in this browser.

https://api.runpod.ai/v2/<ENDPOINT_ID>/openai/v1

Put your endpoint ID in place of <ENDPOINT_ID>. The API key is the one from step 2.

from openai import OpenAI
client = OpenAI(base_url="https://api.runpod.ai/v2/<ENDPOINT_ID>/openai/v1", api_key="rpa_...")
r = client.chat.completions.create(
    model="any",
    messages=[{"role": "user", "content": "Hello!"}])
print(r.choices[0].message.content)

Works with any OpenAI client.

https://api.runpod.ai/v2/<ENDPOINT_ID>/openai/v1

Put your endpoint ID in place of <ENDPOINT_ID>. The API key is the one from step 2.

Free on Windows, macOS, Linux, iOS, Android and the web.

Open the web app ↗ (opens in a new tab)Download ↗ (opens in a new tab)

1. Settings → Model Provider → Add → Add Custom Provider
2. API Mode OpenAI API Compatible; paste the API key and the base URL as API Host
3. Model → Fetch, pick the model
4. New chat → pick the provider → start chatting

Free and open source for Windows, macOS and Linux.

Download Jan ↗ (opens in a new tab)

1.Settings → next to Model Providers press +
2.API format OpenAI-compatible; Base URL is the base URL above (with /v1); API key is your key; Create
3.If no model appears, click + in Models and type any
4.New chat → pick the model → start chatting

How Much VRAM Does Qwen2.5 Coder 7B Instruct Need?

Select a quantization level; the listed hardware adheres to it.

VRAM	+32K	+128K	File	Quality	On RTX 4090 (reference)
2.5 GB	4.3 GB	—	2.3 GB	2-bit · i-quant extreme compression — Importance-weighted 2-bit, extreme compression — significant quality loss	Fits: 32K context · ~261 tok/s
2.7 GB	4.5 GB	—	2.5 GB	2-bit · i-quant extra small — Importance-weighted 2-bit, extra small	Fits: 32K context · ~243 tok/s
2.8 GB	4.6 GB	—	2.6 GB	2-bit · i-quant small — Importance-weighted 2-bit, small	Fits: 32K context · ~234 tok/s
3.0 GB	4.8 GB	—	2.8 GB	2-bit · i-quant medium — Importance-weighted 2-bit, medium	Fits: 32K context · ~219 tok/s
3.4 GB	5.1 GB	—	3.1 GB	3-bit · i-quant — Importance-weighted 3-bit	Fits: 32K context · ~194 tok/s
3.6 GB	5.3 GB	—	3.3 GB	3-bit · i-quant extra small — Importance-weighted 3-bit, extra small	Fits: 32K context · ~184 tok/s
3.6 GB	5.4 GB	—	3 GB	2-bit · K-quant — 2-bit quantization with K-quant improvements	Fits: 32K context · ~180 tok/s
3.6 GB	5.4 GB	—	3.5 GB	3-bit · i-quant small — Importance-weighted 3-bit, small	Fits: 32K context · ~180 tok/s
3.8 GB	5.5 GB	—	3.5 GB	3-bit small — 3-bit small quantization	Fits: 32K context · ~175 tok/s
3.8 GB	5.6 GB	—	3.6 GB	3-bit · i-quant medium — Importance-weighted 3-bit, medium	Fits: 32K context · ~171 tok/s
4.1 GB	5.9 GB	—	3.8 GB	3-bit medium — 3-bit medium quantization	Fits: 32K context · ~159 tok/s
4.2 GB	6 GB	—	4.4 GB	4-bit · legacy — 4-bit legacy quantization	Fits: 32K context · ~155 tok/s
4.3 GB	6.1 GB	—	4.1 GB	3-bit large — 3-bit large quantization	Fits: 32K context · ~152 tok/s
4.5 GB	6.3 GB	—	4.2 GB	4-bit · i-quant compact — Importance-weighted 4-bit, compact	Fits: 32K context · ~145 tok/s
4.7 GB	6.5 GB	—	4.5 GB	4-bit small — 4-bit small quantization	Fits: 32K context · ~139 tok/s
5.0 GB	6.8 GB	—	4.7 GB	4-bit medium — 4-bit medium quantization — most popular sweet spot	Fits: 32K context · ~131 tok/s
5.1 GB	6.8 GB	—	5.1 GB	4-bit large — 4-bit large quantization	Fits: 32K context · ~129 tok/s
5.2 GB	6.9 GB	—	5.3 GB	5-bit · legacy — 5-bit legacy quantization	Fits: 32K context · ~127 tok/s
5.7 GB	7.4 GB	—	5.3 GB	5-bit small — 5-bit small quantization	Fits: 32K context · ~116 tok/s
5.8 GB	7.6 GB	—	5.4 GB	5-bit medium — 5-bit medium quantization — good quality/size tradeoff	Fits: 32K context · ~112 tok/s
5.9 GB	7.7 GB	—	5.8 GB	5-bit large — 5-bit large quantization	Fits: 32K context · ~110 tok/s
6.7 GB	8.5 GB	—	6.3 GB	6-bit, very good — 6-bit quantization, very good quality	Fits: 32K context · ~98 tok/s
8.0 GB	9.8 GB	—	8.1 GB	8-bit, near-lossless — 8-bit quantization, near-lossless	Fits: 32K context · ~82 tok/s
15.7 GB	17.4 GB	—	~15.2 GB	16-bit · bfloat — Brain floating point 16 — preferred for training	Fits: 32K context · ~42 tok/s

est.calculated, no published file yet

Selected Q4_K_M

4-bit medium quantization

5.8GBat 16K context

Weights: 4.6 GB
KV cache: 0.9 GB
Overhead: 0.3 GB

Fits RTX 4090 (reference) · 18.2 GB free · ~131 tok/s

≥50 tok/s needs ≥383 GB/s · RTX 5090 ~233, RTX 3090 Ti ~131

Which Hardware Can Run Qwen2.5 Coder 7B Instruct?

Q4_K_M · 5.0 GB

Qwen2.5 Coder 7B Instruct (Q4_K_M) requires 5.0 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 7+ GB is recommended. Using the full 33K context window can add up to 1.8 GB, bringing total usage to 6.8 GB. 52 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti. 151 devices with unified memory can run Qwen2.5 Coder 7B Instruct, including Mac Studio (M3 Ultra, 256GB), Mac Studio (M3 Ultra, 512GB).

By VRAM tier · typical GPUs and Macs at Q4_K_M

Typical hardwareQuantVRAMtok/sContext that fits

8 GB · RTX 4060Q4_K_M · 5.8 GB · ~35 tok/s · 32K5.8~3532K8 GB tier →

12 GB · RTX 3060 12GBQ4_K_M · 5.8 GB · ~47 tok/s · 32K5.8~4732K12 GB tier →

16 GB · RTX 5060 Ti 16GBQ4_K_M · 5.8 GB · ~58 tok/s · 32K5.8~5832K16 GB tier →

24 GB · RTX 4090referenceQ4_K_M · 5.8 GB · ~131 tok/s · 32K5.8~13132K24 GB tier →

Mac 32 GB · Mac Mini M4Q4_K_M · 5.8 GB · ~17 tok/s · 32K5.8~1732KMacs →

Mac 64 GB · Mac Studio M4 MaxQ4_K_M · 5.8 GB · ~77 tok/s · 32K5.8~7732KMacs →

How good is Qwen2.5 Coder 7B Instruct?

llmrun Score blends public benchmark boards into one number; 100 = the average open model.

Not enough data yet for an llmrun Score.

A Score needs results on enough independent boards. How the Composite Score is built

Benchmarks

GSM8K
Epoch AI · Sep 2024Lab-reported
86.7%
#8/62 open
Lab-reported

Qwen2.5 Coder 7B Instruct benchmark results

Alternatives to Qwen2.5 Coder 7B Instruct

Qwen 2.5 family · 44 models →

Similar models

6tok/s on RTX 4090

Model	Score	VRAM	tok/s	Why it is similar
ALoDLM 8BAmazon	—	7.6 GB	~119	same size
Ling 3.0 TinyInclusion AI	—	5.5 GB	~364	same size
BitCPM CANN 8BOpenBMB	—	5.6 GB	~127	same size
Hy MT2 7BTencent	—	7.3 GB	~122	same size
Gemma 4 E4BGoogle	—	5.6 GB	~123	same size
Olmo 3 7B InstructAllen AI	—	8.4 GB	~114	same size

Fine-tunes & variants(10)community builds of Qwen2.5 Coder 7B Instruct

Specifications

Publisher: Alibaba
Family: Qwen 2.5
Parameters: 7.6B
Architecture: Qwen2ForCausalLM
Context Length: 32,768 tokens
Vocabulary Size: 152,064
Release Date: 2024-09-17
License: Apache 2.0

Resources

Frequently Asked Questions

How much VRAM does Qwen2.5 Coder 7B Instruct need?

Qwen2.5 Coder 7B Instruct requires 5.0 GB of VRAM at Q4_K_M, or 15.7 GB at BF16. Full 33K context adds up to 1.8 GB (6.8 GB total).

VRAM = Weights + KV Cache + Overhead

Weights = 7.6B × 4.8 bits ÷ 8 = 4.6 GB

KV Cache + Overhead ≈ 0.4 GB (at 2K context + ~0.3 GB framework)

Fit ratings and hardware model lists check this model with room for a 16K-token context, which needs a little more memory.

KV Cache + Overhead ≈ 2.2 GB (at full 33K context)

VRAM usage by quantization

Q4_K_M

5.0 GB

Q4_K_M + full context

6.8 GB

Learn more about VRAM estimation →

What's the best quantization for Qwen2.5 Coder 7B Instruct?

For Qwen2.5 Coder 7B Instruct, Q4_K_M (5.0 GB) offers the best balance of quality and VRAM usage. Q4_K_L (5.1 GB) provides better quality if you have the VRAM. The smallest option is IQ2_XXS at 2.5 GB.

VRAM requirement by quantization

IQ2_XXS

2.5 GB~53%

Q2_K

3.6 GB~75%

Q3_K_L

4.3 GB~86%

Q4_K_M ★

5.0 GB~89%

Q5_0

5.2 GB~90%

BF16

15.7 GB~100%

★ Recommended — best balance of quality and VRAM usage.

Learn more about quantization →

Can I run Qwen2.5 Coder 7B Instruct on a Mac?

Yes — MacBook Air 13" M3 (8 GB) and 145 other Macs can run Qwen2.5 Coder 7B Instruct, the smallest at Q2_K. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 5.0 GB of usable unified memory (RAM minus macOS overhead), from Mac Mini M2 Pro (16 GB).

Can I run Qwen2.5 Coder 7B Instruct locally?

Yes — Qwen2.5 Coder 7B Instruct can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.

How fast is Qwen2.5 Coder 7B Instruct?

At Q4_K_M, Qwen2.5 Coder 7B Instruct can reach ~962 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~131 tok/s. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.

tok/s = (bandwidth GB/s ÷ model GB) × efficiency

Example: NVIDIA B200 → 8000 ÷ 4.99 × 0.65 = ~1042 tok/s

Estimated speed at Q4_K_M (5.0 GB)

NVIDIA B200

~1042 tok/s

NVIDIA GeForce RTX 4090

~131 tok/s

NVIDIA B300

~1042 tok/s

AMD Instinct MI350X

~962 tok/s

Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.

Learn more about tok/s estimation →

What's the download size of Qwen2.5 Coder 7B Instruct?

At Q4_K_M, the download is about 4.57 GB. The full-precision BF16 version is 15.23 GB. The smallest option (IQ2_XXS) is 2.09 GB.

Which GPUs can run Qwen2.5 Coder 7B Instruct?

52 consumer GPUs can run Qwen2.5 Coder 7B Instruct at Q4_K_M. Top options include AMD Radeon RX 6700 XT, AMD Radeon RX 6800, AMD Radeon RX 6800 XT. 52 GPUs have plenty of headroom for comfortable inference.

Which devices can run Qwen2.5 Coder 7B Instruct?

151 devices with unified memory can run Qwen2.5 Coder 7B Instruct at Q4_K_M (5.0 GB), including AMD Ryzen AI 9 HX 370 (Strix Point) Laptop, ASUS Ascent GX10, Beelink GTR9 Pro (Ryzen AI Max+ 395, 128 GB), Framework Desktop (Ryzen AI Max+ 395, 128 GB). Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.