Can the NVIDIA L40 run PaddleOCR-VL 1.6 0.9B?

Yes, it runs great

Runs great. At F16 it needs about 5.1 GB of 48.0 GB usable, with an estimated speed of 259 tok/s.

Best quant: F16
Generation speed: 259 tok/s
Prompt processing: 67,875 tok/s
First token (1K prompt): 0.3 s
Largest safe context: 128K

Best app to run it on the NVIDIA L40

On Linux

Inference server · runs on vLLM

The full model fits in GPU memory at 8-bit, so it can batch many requests: best for serving a team or an app.

vllm serve PaddlePaddle/PaddleOCR-VL-1.6

LM Studio Point-and-click: search, download and chat in one window.
SGLang The full model fits in GPU memory at 8-bit, so it can batch many requests: best for serving a team or an app.

On Windows

LM Studio

Desktop app · runs on llama.cpp, MLX

Point-and-click: search, download and chat in one window.

Open PaddleOCR-VL 1.6 0.9B in LM Studio

llama.cpp Usually the first to support new model architectures.
Ollama Runs the Hugging Face GGUF with ollama run hf.co/…

How we choose: the files published for this model, your operating system and GPU, whether it fits fully in memory, and published speed tests (MLX vs llama.cpp on Apple silicon, vLLM vs llama.cpp vs Ollama on NVIDIA). Details.

Download PaddleOCR-VL 1.6 0.9B

Quantized files · GGUF

Works in LM Studio, llama.cpp, Jan and Ollama. No sign-in needed.

huggingface.co/PaddlePaddle/PaddleOCR-VL-1.6-GGUF (opens in a new tab)

ollama run hf.co/PaddlePaddle/PaddleOCR-VL-1.6-GGUF

Original weights · Hugging Face

Full-precision files from Baidu, for fine-tuning, vLLM, Transformers.

huggingface.co/PaddlePaddle/PaddleOCR-VL-1.6 (opens in a new tab)

No sign-in needed.

Official repo · GitHub

Reference code, inference scripts and docs from Baidu. The weights themselves download from Hugging Face.

github.com/PaddlePaddle/PaddleOCR (opens in a new tab)

git clone https://github.com/PaddlePaddle/PaddleOCR

Links checked 3 days ago against Hugging Face, Ollama and GitHub.

How it’s rated

Quality 42Fair quality (C) · 42 of 100: how capable the model is at F16, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking.: How capable the model itself is at F16, from its editorial quality tier scaled by the quantization or precision. It’s the same on any hardware that runs this variant. Scale: A+ 85+ · A 70+ · B 55+ · C 40+ · D under 40.
Runs Runs great: How well it runs on the NVIDIA L40: speed, capped by how well it fits in memory. Runs great: 30+ tokens/s (or near the image, video or decision speed target) · Runs well: 10+ tokens/s · Slow: 3+ tokens/s · Very slow: under 3 · Won’t run: doesn’t fit even with offload. A tight fit or RAM offload caps it at Runs well; heavy offload caps it at Slow.

Match score

To order lists, each result also gets a match score out of 100 from four parts. It favors the most capable model that still runs well.

Model qualityEditorial quality tier of the model, scaled by the quantization or precision used.	18.9 / 45
Memory fitFull marks when everything sits in GPU memory; fewer when part spills to system RAM.	25.0 / 25
SpeedGeneration speed against a 40 tokens/s target for LLMs, or render time against a target for image and video.	15.0 / 15
Memory useRewards using the memory you have, so a bigger, better variant beats a tiny one that leaves most memory idle.	2.4 / 15
Match score	61.3 / 100

The match score orders results; it is not shown as a grade. Quality and Runs are the two ratings in the tables.

Every quantization on the NVIDIA L40

Quant	Memory needed	Available	Speed	Runs	Quality	Download
`Q2_K`	3.5 GB	48.0 GB	1,814 tok/s	Runs great	29Basic quality (D) · 29 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking.	GGUF files (opens in a new tab)
`Q3_K_M`	3.6 GB	48.0 GB	1,428 tok/s	Runs great	35Basic quality (D) · 35 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking.	GGUF files (opens in a new tab)
`Q4_K_M`	3.7 GB	48.0 GB	1,105 tok/s	Runs great	39Basic quality (D) · 39 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking.	GGUF files (opens in a new tab)
`Q5_K_M`	3.8 GB	48.0 GB	913 tok/s	Runs great	40Fair quality (C) · 40 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking.	GGUF files (opens in a new tab)
`Q6_K`	3.9 GB	48.0 GB	767 tok/s	Runs great	41Fair quality (C) · 41 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking.	GGUF files (opens in a new tab)
`Q8_0`	4.2 GB	48.0 GB	562 tok/s	Runs great	42Fair quality (C) · 42 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking.	GGUF files (opens in a new tab)
`F16`	5.1 GB	48.0 GB	259 tok/s	Runs great	42Fair quality (C) · 42 of 100: how capable the model is at F16, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking.	Original weights (opens in a new tab)

Memory figures assume 8K of context. Longer conversations or documents grow the KV cache.

Other hardware that runs it

42Fair quality (C) · 42 of 100: how capable the model is at F16, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. NVIDIA GeForce RTX 4060 F16 · 81.6 tok/s · Runs great
42Fair quality (C) · 42 of 100: how capable the model is at F16, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. Intel Arc B580 F16 · 105 tok/s · Runs great
42Fair quality (C) · 42 of 100: how capable the model is at F16, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. NVIDIA GeForce RTX 3060 12GB F16 · 108 tok/s · Runs great
42Fair quality (C) · 42 of 100: how capable the model is at F16, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. NVIDIA GeForce RTX 4070 SUPER F16 · 151 tok/s · Runs great
42Fair quality (C) · 42 of 100: how capable the model is at F16, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. Apple M4 Pro (24 GB) F16 · 70.5 tok/s · Runs great
42Fair quality (C) · 42 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. Apple M4 (16 GB) Q8_0 · 67.3 tok/s · Runs great

Quick answers

How much memory does PaddleOCR-VL 1.6 0.9B need on the NVIDIA L40 48GB?

At F16 it needs about 5.1 GB. The NVIDIA L40 48GB has 48.0 GB usable.

How fast is PaddleOCR-VL 1.6 0.9B on the NVIDIA L40 48GB?

About 259 tok/s at F16, an estimate from the hardware’s memory bandwidth and the size of the model.

Which quantization of PaddleOCR-VL 1.6 0.9B should I use on the NVIDIA L40 48GB?

F16. Of the versions that fit, it is the one that runs best; when two run equally well, the check prefers the higher-rated one.

Estimates come from published hardware specs and our open formulas, not from running the model on your machine. Real speed varies with runtime, drivers and settings. Numbers marked “Measured” come from a linked benchmark.