BigScienceBloomForCausalLM

Bloom 560M — Hardware Requirements & GPU Compatibility

Chat

Bloom 560M is a 559M-parameter open language model from BigScience. At Q4_K_M it needs about 0.37 GB of VRAM — see which GPUs and Macs can run it below.

559M params · densebigscience-bloom-rail-1.0May 2022377.6K downloads376 likes1.8K 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.

S

Runs great on a 24 GB GPU

Runs from 8 GB (Q4_K_M · 32K)Best on 24 GB (Q4_K_M · ~1771 tok/s · 32K)Macs from 8 GB (Q4_K_M · ~194 tok/s)

bigscience-bloom-rail-1.0

Reference GPU · RTX 4090 24 GB

Estimates, ±20% · how we estimate

Run Bloom 560M

One recommended way. Every other method, OS and quant is in the step-by-step guide.

Recommended

Ollama

one command · needs Ollama installed

$ollama run hf.co/afrideva/bloom-560m-GGUF:Q4_K_M

Q4_K_M fits fully on the RTX 4090 · 22.4 GB free · 16K context

Download: bloom-560m.q4_k_m.gguf0.56 GBafrideva

install, OS, other quants

Other ways

each opens step-by-step

Run in cloud

Fits on RTX 3060 12GB (10 GB headroom) · Q4_K_M

Generation speed: ~632 tok/s; generation speed
Cost per 1M output tokens: $0.036; per 1M output tokens

Deployon Vast.aiBest optionfrom $0.083/hr

Compare GPUs →

or

How Much VRAM Does Bloom 560M Need?

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

Quantization	VRAM	+32K	+128K	File	Quality	On RTX 4090 (reference)
	0.2 GB	2.8 GB	11.1 GB	~0.2 GB	2-bit · i-quant extreme compression — Importance-weighted 2-bit, extreme compression — significant quality loss	Fits: 32K context · ~3854 tok/s
	0.2 GB	2.8 GB	11.1 GB	~0.2 GB	2-bit · i-quant extra small — Importance-weighted 2-bit, extra small	Fits: 32K context · ~3640 tok/s
	0.2 GB	2.8 GB	11.2 GB	~0.2 GB	2-bit · i-quant small — Importance-weighted 2-bit, small	Fits: 32K context · ~3448 tok/s
	0.2 GB	2.8 GB	11.2 GB	~0.2 GB	2-bit · i-quant medium — Importance-weighted 2-bit, medium	Fits: 32K context · ~3120 tok/s
	0.2 GB	2.9 GB	11.2 GB	~0.2 GB	3-bit · i-quant — Importance-weighted 3-bit	Fits: 32K context · ~2730 tok/s
	0.3 GB	2.9 GB	11.2 GB	~0.2 GB	2-bit · small K-quant — 2-bit small K-quant	Fits: 32K context · ~2621 tok/s
	0.3 GB	2.9 GB	11.2 GB	~0.2 GB	3-bit · i-quant extra small — Importance-weighted 3-bit, extra small	Fits: 32K context · ~2621 tok/s
	0.3 GB	2.9 GB	11.2 GB	~0.2 GB	3-bit · i-quant small — Importance-weighted 3-bit, small	Fits: 32K context · ~2520 tok/s
	0.3 GB	2.9 GB	11.2 GB	~0.2 GB	2-bit · K-quant — 2-bit quantization with K-quant improvements	Fits: 32K context · ~2520 tok/s
	0.3 GB	2.9 GB	11.2 GB	~0.2 GB	3-bit small — 3-bit small quantization	Fits: 32K context · ~2427 tok/s
	0.3 GB	2.9 GB	11.2 GB	~0.3 GB	3-bit · i-quant medium — Importance-weighted 3-bit, medium	Fits: 32K context · ~2340 tok/s
	0.3 GB	2.9 GB	11.3 GB	~0.3 GB	3-bit medium — 3-bit medium quantization	Fits: 32K context · ~2184 tok/s
	0.3 GB	2.9 GB	11.3 GB	~0.3 GB	4-bit · legacy — 4-bit legacy quantization	Fits: 32K context · ~2114 tok/s
	0.3 GB	2.9 GB	11.3 GB	~0.3 GB	3-bit large — 3-bit large quantization	Fits: 32K context · ~2048 tok/s
	0.3 GB	2.9 GB	11.3 GB	~0.3 GB	4-bit · i-quant compact — Importance-weighted 4-bit, compact	Fits: 32K context · ~1986 tok/s
	0.3 GB	3 GB	11.3 GB	~0.3 GB	4-bit · legacy+ — 4-bit legacy quantization with offset	Fits: 32K context · ~1872 tok/s
	0.3 GB	3 GB	11.3 GB	~0.3 GB	4-bit small — 4-bit small quantization	Fits: 32K context · ~1872 tok/s
	0.3 GB	3 GB	11.3 GB	~0.3 GB	4-bit · i-quant non-linear — Importance-weighted 4-bit, non-linear	Fits: 32K context · ~1872 tok/s
	0.4 GB	3 GB	11.3 GB	~0.3 GB	4-bit medium — 4-bit medium quantization — most popular sweet spot	Fits: 32K context · ~1771 tok/s
	0.4 GB	3 GB	11.4 GB	~0.4 GB	5-bit small — 5-bit small quantization	Fits: 32K context · ~1560 tok/s
	0.4 GB	3.1 GB	11.4 GB	~0.4 GB	5-bit medium — 5-bit medium quantization — good quality/size tradeoff	Fits: 32K context · ~1489 tok/s
	0.5 GB	3.1 GB	11.5 GB	0.7 GB	6-bit, very good — 6-bit quantization, very good quality	Fits: 32K context · ~1285 tok/s
	0.6 GB	3.2 GB	11.6 GB	~0.6 GB	8-bit, near-lossless — 8-bit quantization, near-lossless	Fits: 32K context · ~1057 tok/s
	1.2 GB	3.8 GB	12.2 GB	~1.1 GB	16-bit · bfloat — Brain floating point 16 — preferred for training	Fits: 32K context · ~533 tok/s

est.calculated, no published file yet

Selected Q4_K_M

4-bit medium quantization

1.6GBat 16K context

Weights: 0.1 GB
KV cache: 1.2 GB
Overhead: 0.3 GB

Fits RTX 4090 (reference) · 22.4 GB free · ~1771 tok/s

≥50 tok/s needs ≥28 GB/s · RTX 5090 ~3148, RTX 3090 Ti ~1771

Which Hardware Can Run Bloom 560M?

Q4_K_M · 0.4 GB

Bloom 560M (Q4_K_M) requires 0.4 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 Bloom 560M, 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 · 1.6 GB · ~478 tok/s · 32K1.6~47832K8 GB tier →

12 GB · RTX 3060 12GBQ4_K_M · 1.6 GB · ~632 tok/s · 32K1.6~63232K12 GB tier →

16 GB · RTX 5060 Ti 16GBQ4_K_M · 1.6 GB · ~787 tok/s · 32K1.6~78732K16 GB tier →

24 GB · RTX 4090referenceQ4_K_M · 1.6 GB · ~1771 tok/s · 32K1.6~177132K24 GB tier →

Mac 32 GB · Mac Mini M4Q4_K_M · 1.6 GB · ~227 tok/s · 32K1.6~22732KMacs →

Mac 64 GB · Mac Studio M4 MaxQ4_K_M · 1.6 GB · ~1033 tok/s · 32K1.6~103332KMacs →

Alternatives to Bloom 560M

Similar models

4tok/s on RTX 4090

Model	Score	VRAM	tok/s	Why it is similar
T5gemma B B Ul2 ITGoogle	—	2.5 GB	~504	same size
Falcon H1 0.5B BaseTII UAE	—	1.2 GB	~950	same size
Qwen3 0.6B BaseAlibaba	—	1.6 GB	~840	same size
Xglm 564MMeta	—	0.4 GB	~1771	same size

Fine-tunes & variants(1)community builds of Bloom 560M

Bigscience Bloom 560M

Specifications

Publisher: BigScience
Parameters: 559M
Architecture: BloomForCausalLM
Vocabulary Size: 250,880
Release Date: 2022-05-19
License: bigscience-bloom-rail-1.0

Resources

Hugging Face model card (opens in a new tab)

Frequently Asked Questions

How much VRAM does Bloom 560M need?

Bloom 560M requires 0.4 GB of VRAM at Q4_K_M, or 1.2 GB at BF16.

VRAM = Weights + KV Cache + Overhead

Weights = 559M × 4.8 bits ÷ 8 = 0.3 GB

KV Cache + Overhead ≈ 0.1 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.

VRAM usage by quantization

Q4_K_M

0.4 GB

Learn more about VRAM estimation →

What's the best quantization for Bloom 560M?

For Bloom 560M, Q4_K_M (0.4 GB) offers the best balance of quality and VRAM usage. Q5_K_S (0.4 GB) provides better quality if you have the VRAM. The smallest option is IQ2_XXS at 0.2 GB.

VRAM requirement by quantization

IQ2_XXS

0.2 GB~53%

IQ3_XS

0.3 GB~73%

Q4_0

0.3 GB~85%

IQ4_NL

0.3 GB~88%

Q4_K_M ★

0.4 GB~89%

BF16

1.2 GB~100%

★ Recommended — best balance of quality and VRAM usage.

Learn more about quantization →

Can I run Bloom 560M on a Mac?

Yes — MacBook Air 13" M3 (8 GB) and 145 other Macs can run Bloom 560M. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 0.4 GB of usable unified memory (RAM minus macOS overhead).

Can I run Bloom 560M locally?

Yes — Bloom 560M can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.

How fast is Bloom 560M?

At Q4_K_M, Bloom 560M can reach ~12973 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~1771 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 ÷ 0.37 × 0.65 = ~14054 tok/s

Estimated speed at Q4_K_M (0.4 GB)

NVIDIA B200

~14054 tok/s

NVIDIA GeForce RTX 4090

~1771 tok/s

NVIDIA B300

~14054 tok/s

AMD Instinct MI350X

~12973 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 Bloom 560M?

At Q4_K_M, the download is about 0.34 GB. The full-precision BF16 version is 1.12 GB. The smallest option (IQ2_XXS) is 0.15 GB.

Which GPUs can run Bloom 560M?

52 consumer GPUs can run Bloom 560M 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 Bloom 560M?

163 devices with unified memory can run Bloom 560M at Q4_K_M (0.4 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.

Run Bloom 560M

1.6 GB VRAM at 16K context · fits the RTX 4090 (22.4 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/afrideva/bloom-560m-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/afrideva/bloom-560m-GGUF, then pick Q4_K_M. Or use the CLI (run LM Studio once first):
$lms get https://huggingface.co/afrideva/bloom-560m-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 afrideva/bloom-560m-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/afrideva/bloom-560m-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 afrideva/bloom-560m-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 bigscience/bloom-560m --max-model-len 16384
Runs the original weights, not the Q4_K_M GGUF; needs ~1 GB of GPU memory.

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.
$afrideva/bloom-560m-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 3060 12GBCheapestCheapest10 GB free	~632 tok/s	$0.036	Deploy on Vast.aifrom $0.083/hr (opens in new tab)	10 GB free · ~632 tok/s · $0.036 / 1M tokens
RTX 3070 (8 GB)6 GB free	~787 tok/s	$0.046	Deploy on RunPodfrom $0.13/hr (opens in new tab)	8 GB · 6 GB free · ~787 tok/s · $0.046 / 1M tokens
RTX 3070 (8 GB)6 GB free	~787 tok/s	$0.048	Deploy on Vast.aifrom $0.14/hr (opens in new tab)	8 GB · 6 GB free · ~787 tok/s · $0.048 / 1M tokens
RTX 3060 Ti (8 GB)6 GB free	~787 tok/s	$0.053	Deploy on Vast.aifrom $0.15/hr (opens in new tab)	8 GB · 6 GB free · ~787 tok/s · $0.053 / 1M tokens
RTX 5060 Ti 16GB14 GB free	~787 tok/s	$0.058	Deploy on Vast.aifrom $0.16/hr (opens in new tab)	14 GB free · ~787 tok/s · $0.058 / 1M tokens
RTX 3080 Ti (12 GB)10 GB free	~1603 tok/s	$0.031	Deploy on RunPodfrom $0.18/hr (opens in new tab)	12 GB · 10 GB free · ~1603 tok/s · $0.031 / 1M tokens
Tesla V100 PCIe 16GB14 GB free	~1581 tok/s	$0.033	Deploy on RunPodfrom $0.19/hr (opens in new tab)	14 GB free · ~1581 tok/s · $0.033 / 1M tokens
RTX 4090 (24 GB)22 GB free	~1771 tok/s	$0.053	Deploy on RunPodfrom $0.34/hr (opens in new tab)	24 GB · 22 GB free · ~1771 tok/s · $0.053 / 1M tokens

Compare GPUs →