OpenAIGPT2LMHeadModel

Gpt2 — Hardware Requirements & GPU Compatibility

Chat

GPT-2 is the landmark 2019 language model from OpenAI that helped ignite widespread interest in large-scale text generation. At only 137 million parameters it is tiny by modern standards, but it holds an important place in AI history as the model that was initially deemed too dangerous to release in full. Today GPT-2 runs effortlessly on virtually any hardware, including CPUs, making it ideal for educational purposes, experimentation, and understanding transformer fundamentals. It should not be expected to match the quality of modern instruction-tuned models, but it remains a useful teaching tool and conversation starter.

137M params · dense1K contextMITMar 202230.8M downloads3.9K likes8.7K quant downloads

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)Best on 24 GB (Q4_K_M · ~7280 tok/s)Macs from 8 GB (Q4_K_M · ~796 tok/s)

MIT

Reference GPU · RTX 4090 24 GB

Estimates, ±20% · how we estimate

Run Gpt2

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

Runs on RTX 3080 Ti (12 GB) (11 GB headroom) · Q4_K_M

Generation speed: ~6590 tok/s; generation speed
Cost per 1M output tokens: $0.0052; per 1M output tokens

Deployon Vast.aiBest valuefrom $0.12/hr

Compare GPUs →

Run Gpt2

0.1 GB VRAM at 1K context · fits the RTX 4090 (23.9 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 hf.co/mradermacher/gpt2-GGUF:Q4_K_M
Raise the context window
Type it inside the chat. For the API server instead, start it with OLLAMA_CONTEXT_LENGTH=1024 ollama serve.
$/set parameter num_ctx 1024
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/mradermacher/gpt2-GGUF, then pick Q4_K_M. Or use the CLI (run LM Studio once first):
$lms get https://huggingface.co/mradermacher/gpt2-GGUF@Q4_K_M
Open on Hugging Face ↗
Load and chat
Open the model loader, select the model and set Context Length to 1024. 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 mradermacher/gpt2-GGUF:Q4_K_M -c 1024 -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/mradermacher/gpt2-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 mradermacher/gpt2-GGUF:Q4_K_M -c 1024 -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 openai-community/gpt2 --max-model-len 1024
Runs the original weights, not the Q4_K_M GGUF; needs ~0 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 jesusoctavioas/gpt2-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 jesusoctavioas/gpt2-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.
$mradermacher/gpt2-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)CheapestCheapest11 GB free	~6590 tok/s	$0.0052	Deploy on Vast.aifrom $0.12/hr (opens in new tab)	12 GB · 11 GB free · ~6590 tok/s · $0.0052 / 1M tokens
RTX 3060 12GB11 GB free	~2600 tok/s	$0.0066	Deploy on Vast.aifrom $0.061/hr (opens in new tab)	11 GB free · ~2600 tok/s · $0.0066 / 1M tokens
RTX 4060 (8 GB)7 GB free	~1964 tok/s	$0.012	Deploy on Vast.aifrom $0.086/hr (opens in new tab)	8 GB · 7 GB free · ~1964 tok/s · $0.012 / 1M tokens
RTX A4000 (16 GB)15 GB free	~3236 tok/s	$0.0076	Deploy on Vast.aifrom $0.088/hr (opens in new tab)	16 GB · 15 GB free · ~3236 tok/s · $0.0076 / 1M tokens
RTX 4070 SUPER (12 GB)11 GB free	~3640 tok/s	$0.0079	Deploy on Vast.aifrom $0.10/hr (opens in new tab)	12 GB · 11 GB free · ~3640 tok/s · $0.0079 / 1M tokens
RTX PRO 4500 (32 GB)31 GB free	~6471 tok/s	$0.015	Deploy on RunPodfrom $0.34/hr (opens in new tab)	32 GB · 31 GB free · ~6471 tok/s · $0.015 / 1M tokens
RTX 4090 (24 GB)23 GB free	~7280 tok/s	$0.013	Deploy on RunPodfrom $0.34/hr (opens in new tab)	24 GB · 23 GB free · ~7280 tok/s · $0.013 / 1M tokens
A40 (48 GB)47 GB free	~5027 tok/s	$0.019	Deploy on RunPodfrom $0.35/hr (opens in new tab)	48 GB · 47 GB free · ~5027 tok/s · $0.019 / 1M tokens
L4 (24 GB)23 GB free	~2167 tok/s	$0.056	Deploy on RunPodfrom $0.44/hr (opens in new tab)	24 GB · 23 GB free · ~2167 tok/s · $0.056 / 1M tokens

Compare GPUs →

How Much VRAM Does Gpt2 Need?

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

VRAM	+32K	+128K	File	Quality	On RTX 4090 (reference)
0.0 GB	—	—	~0 GB	2-bit · i-quant extreme compression — Importance-weighted 2-bit, extreme compression — significant quality loss	Fits: — context · ~16380 tok/s
0.1 GB	—	—	~0 GB	2-bit · i-quant extra small — Importance-weighted 2-bit, extra small	Fits: — context · ~13104 tok/s
0.1 GB	—	—	~0.1 GB	2-bit · i-quant medium — Importance-weighted 2-bit, medium	Fits: — context · ~13104 tok/s
0.1 GB	—	—	~0 GB	2-bit · i-quant small — Importance-weighted 2-bit, small	Fits: — context · ~13104 tok/s
0.1 GB	—	—	0.1 GB	3-bit · i-quant extra small — Importance-weighted 3-bit, extra small	Fits: — context · ~10920 tok/s
0.1 GB	—	—	~0.1 GB	3-bit · i-quant — Importance-weighted 3-bit	Fits: — context · ~10920 tok/s
0.1 GB	—	—	0.1 GB	2-bit · K-quant — 2-bit quantization with K-quant improvements	Fits: — context · ~10920 tok/s
0.1 GB	—	—	~0.1 GB	3-bit · i-quant small — Importance-weighted 3-bit, small	Fits: — context · ~10920 tok/s
0.1 GB	—	—	~0.1 GB	3-bit · i-quant medium — Importance-weighted 3-bit, medium	Fits: — context · ~9360 tok/s
0.1 GB	—	—	~0.1 GB	3-bit small — 3-bit small quantization	Fits: — context · ~9360 tok/s
0.1 GB	—	—	0.1 GB	3-bit medium — 3-bit medium quantization	Fits: — context · ~9360 tok/s
0.1 GB	—	—	~0.1 GB	4-bit · legacy — 4-bit legacy quantization	Fits: — context · ~8190 tok/s
0.1 GB	—	—	0.1 GB	4-bit small — 4-bit small quantization	Fits: — context · ~8190 tok/s
0.1 GB	—	—	0.1 GB	3-bit large — 3-bit large quantization	Fits: — context · ~8190 tok/s
0.1 GB	—	—	0.1 GB	4-bit · legacy+ — 4-bit legacy quantization with offset	Fits: — context · ~8190 tok/s
0.1 GB	—	—	0.1 GB	4-bit · i-quant compact — Importance-weighted 4-bit, compact	Fits: — context · ~8190 tok/s
0.1 GB	—	—	0.1 GB	4-bit medium — 4-bit medium quantization — most popular sweet spot	Fits: — context · ~7280 tok/s
0.1 GB	—	—	0.1 GB	5-bit · legacy — 5-bit legacy quantization	Fits: — context · ~7280 tok/s
0.1 GB	—	—	0.1 GB	5-bit · legacy+ — 5-bit legacy quantization with offset	Fits: — context · ~6552 tok/s
0.1 GB	—	—	0.1 GB	5-bit small — 5-bit small quantization	Fits: — context · ~6552 tok/s
0.1 GB	—	—	0.1 GB	5-bit medium — 5-bit medium quantization — good quality/size tradeoff	Fits: — context · ~5956 tok/s
0.1 GB	—	—	0.1 GB	6-bit, very good — 6-bit quantization, very good quality	Fits: — context · ~5460 tok/s
0.1 GB	—	—	0.2 GB	8-bit, near-lossless — 8-bit quantization, near-lossless	Fits: — context · ~4368 tok/s
0.3 GB	—	—	~0.3 GB	16-bit · bfloat — Brain floating point 16 — preferred for training	Fits: — context · ~2184 tok/s

est.calculated, no published file yet

Selected Q4_K_M

4-bit medium quantization

0.1GBat 1K context

Weights: 0.1 GB
KV cache: 0.0 GB
Overhead: 0.3 GB

Fits RTX 4090 (reference) · 23.9 GB free · ~7280 tok/s

≥50 tok/s needs ≥7 GB/s · RTX 5090 ~12942, RTX 3090 Ti ~7280

Which Hardware Can Run Gpt2?

Q4_K_M · 0.1 GB

Gpt2 (Q4_K_M) requires 0.1 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 1+ GB is recommended. 52 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti. 163 devices with unified memory can run Gpt2, 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 · 0.1 GB · ~1964 tok/s · —0.1~1964—8 GB tier →

12 GB · RTX 3060 12GBQ4_K_M · 0.1 GB · ~2600 tok/s · —0.1~2600—12 GB tier →

16 GB · RTX 5060 Ti 16GBQ4_K_M · 0.1 GB · ~3236 tok/s · —0.1~3236—16 GB tier →

24 GB · RTX 4090referenceQ4_K_M · 0.1 GB · ~7280 tok/s · —0.1~7280—24 GB tier →

Mac 32 GB · Mac Mini M4Q4_K_M · 0.1 GB · ~933 tok/s · —0.1~933—Macs →

Mac 64 GB · Mac Studio M4 MaxQ4_K_M · 0.1 GB · ~4247 tok/s · —0.1~4247—Macs →

Alternatives to Gpt2

Similar models

1tok/s on RTX 4090

Model	Score	VRAM	tok/s	Why it is similar
SmolLM 135M Instruct FP32modularai	—	0.4 GB	~1524	same size

Fine-tunes & variants(1)community builds of Gpt2

DrugGen

Specifications

Publisher: OpenAI
Parameters: 137M
Architecture: GPT2LMHeadModel
Context Length: 1,024 tokens
Vocabulary Size: 50,257
Release Date: 2022-03-02
License: MIT

Resources

Hugging Face model card (opens in a new tab)

Frequently Asked Questions

How much VRAM does Gpt2 need?

Gpt2 requires 0.1 GB of VRAM at Q4_K_M, or 0.3 GB at BF16.

VRAM = Weights + KV Cache + Overhead

Weights = 137M × 4.8 bits ÷ 8 = 0.1 GB

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

VRAM usage by quantization

Q4_K_M

0.1 GB

Learn more about VRAM estimation →

What's the best quantization for Gpt2?

For Gpt2, Q4_K_M (0.1 GB) offers the best balance of quality and VRAM usage. Q5_0 (0.1 GB) provides better quality if you have the VRAM. The smallest option is IQ2_XXS at 0.0 GB.

VRAM requirement by quantization

IQ2_XXS

0.0 GB~53%

Q2_K

0.1 GB~75%

Q4_K_S

0.1 GB~88%

Q4_K_M ★

0.1 GB~89%

Q5_0

0.1 GB~90%

BF16

0.3 GB~100%

★ Recommended — best balance of quality and VRAM usage.

Learn more about quantization →

Can I run Gpt2 on a Mac?

Yes — MacBook Air 13" M3 (8 GB) and 145 other Macs can run Gpt2. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 0.1 GB of usable unified memory (RAM minus macOS overhead).

Can I run Gpt2 locally?

Yes — Gpt2 can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.

How fast is Gpt2?

At Q4_K_M, Gpt2 can reach ~53333 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~7280 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 ÷ 0.09 × 0.65 = ~57778 tok/s

Estimated speed at Q4_K_M (0.1 GB)

NVIDIA B200

~57778 tok/s

NVIDIA GeForce RTX 4090

~7280 tok/s

NVIDIA B300

~57778 tok/s

AMD Instinct MI350X

~53333 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 Gpt2?

At Q4_K_M, the download is about 0.08 GB. The full-precision BF16 version is 0.27 GB. The smallest option (IQ2_XXS) is 0.04 GB.

Which GPUs can run Gpt2?

52 consumer GPUs can run Gpt2 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 Gpt2?

163 devices with unified memory can run Gpt2 at Q4_K_M (0.1 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.