LightOnOCR 2 1B — Hardware Requirements & GPU Compatibility
Vision
LightOnOCR-2-1B is LightOn's flagship end-to-end OCR vision-language model, a 1-billion-parameter system that converts documents such as PDFs, scans, and photos directly into clean, naturally ordered markdown without a separate OCR pipeline. It handles tables, receipts, forms, multi-column layouts, and math notation, is trained on a large multilingual corpus with particular strength in French, arXiv papers, and scanned documents, and this second-generation release adds RLVR (reinforcement learning from verifiable rewards) training on top of the base model for extra accuracy. LightOn reports state-of-the-art results on the OlmOCR-Bench benchmark while being roughly nine times smaller and several times faster than competing OCR systems. At just 1 billion parameters, it runs comfortably even on a modest single consumer GPU. Context length is 16,384 tokens. It is released under the Apache 2.0 license, permitting unrestricted commercial and research use, and was published in January 2026, alongside base, bounding-box, and merged "soup" variants of the same model.
Runs great on a 24 GB GPU
Runs from 8 GB (Q4_K_M · 16K)Best on 24 GB (Q4_K_M · ~642 tok/s · 16K)Macs from 8 GB (Q4_K_M · ~70 tok/s)
Apache 2.0
Reference GPU · RTX 4090 24 GB
Estimates, ±20% · how we estimate
Run LightOnOCR 2 1B
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.
Recommended
ollama run hf.co/wangjazz/LightOnOCR-2-1B-gguf:Q4_K_MQ4_K_M fits fully on the RTX 4090 · 22.2 GB free · 16K context
Download: LightOnOCR-2-1B-Q4_K_M.gguf0.40 GBwangjazz
Ways to runOther ways
each opens step-by-stepRun in cloud
Runs on RTX 3080 Ti (12 GB) (10 GB headroom) · Q4_K_M
- Generation speed
- ~581 tok/s
- generation speed
- Cost per 1M output tokens
- $0.058
- per 1M output tokens
How Much VRAM Does LightOnOCR 2 1B Need?
Select a quantization level; the listed hardware adheres to it.
| Quantization | VRAM | +32K | On RTX 4090 (reference) | |
|---|---|---|---|---|
| 0.8 GB | — | 2-bit · K-quant — 2-bit quantization with K-quant improvements | Fits: 16K context · ~780 tok/s | |
| 0.9 GB | — | 3-bit small — 3-bit small quantization | Fits: 16K context · ~762 tok/s | |
| 0.9 GB | — | 3-bit medium — 3-bit medium quantization | Fits: 16K context · ~720 tok/s | |
| 0.9 GB | — | 4-bit · legacy — 4-bit legacy quantization | Fits: 16K context · ~712 tok/s | |
| 1.0 GB | — | 4-bit medium — 4-bit medium quantization — most popular sweet spot | Fits: 16K context · ~642 tok/s | |
| 1.1 GB | — | 5-bit medium — 5-bit medium quantization — good quality/size tradeoff | Fits: 16K context · ~580 tok/s | |
| 1.3 GB | — | 6-bit, very good — 6-bit quantization, very good quality | Fits: 16K context · ~524 tok/s | |
| 1.4 GB | — | 8-bit, near-lossless — 8-bit quantization, near-lossless | Fits: 16K context · ~461 tok/s |
Selected Q4_K_M
4-bit medium quantization1.8GBat 16K context
- Weights
- 0.6 GB
- KV cache
- 0.9 GB
- Overhead
- 0.3 GB
Fits RTX 4090 (reference) · 22.2 GB free · ~642 tok/s
≥50 tok/s needs ≥78 GB/s · RTX 5090 ~1142, RTX 3090 Ti ~642
Which Hardware Can Run LightOnOCR 2 1B?
Q4_K_M · 1.0 GBLightOnOCR 2 1B (Q4_K_M) requires 1.0 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 2+ GB is recommended. Using the full 16K context window can add up to 0.8 GB, bringing total usage to 1.8 GB. 52 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti. 163 devices with unified memory can run LightOnOCR 2 1B, including Mac Studio (M3 Ultra, 256GB), Mac Studio (M3 Ultra, 512GB).
≤ 11 GB
≤ 11 GB tier →fits · 6.2 GB free · up to 16K ctx
16 GPUs · 1 Mac class · 2 device families
RTX 3080~485Simulate output on RTX 3080RTX 2080 Ti~393Simulate output on RTX 2080 TiRTX 3070 Ti~388Simulate output on RTX 3070 TiM3*8 GB~70Simulate output on iMac 24" M3 (8 GB)Orin Nano~65Simulate output on NVIDIA Jetson Orin Nano 8GB (Super)iPhone~47Simulate output on iPhone 1720 GB
fits · 18.2 GB free · up to 16K ctx
1 GPU
RX 7900 XT~471Simulate output on RX 7900 XT32–40 GB
fits · 30.2 GB free · up to 16K ctx
1 GPU · 3 Mac classes · 4 device families
RTX 5090~1142Simulate output on RTX 5090M5 Max36 GB~316Simulate output on Mac Studio M5 Max (36 GB)M532 GB~105Simulate output on MacBook Air 13" M5 (32 GB)Orin~131Simulate output on NVIDIA Jetson AGX Orin 32GBSnapdragon X Elite (X1E-80-100)~86Simulate output on Snapdragon X Elite Copilot+ PC
By VRAM tier · typical GPUs and Macs at Q4_K_M
Similar models
5| Model | Score | VRAM | tok/s | Why it is similar |
|---|---|---|---|---|
| MiniCPM5 1BOpenBMB | — | 1.3 GB | ~662 | same size |
| Gemma 3 1B ITGoogle | — | 2.1 GB | ~993 | same size |
| LLaMA3.1 8B Instruct DFlash UltraChatz-lab | — | 2.7 GB | ~269 | same size |
| Vaultgemma 1BGoogle | — | 3.7 GB | ~286 | same size |
| Qwen3 8B Speculator.eagle3RedHatAI | — | 2.1 GB | ~978 | same size |
Specifications
- Publisher
- lightonai
- Parameters
- 1.0B
- Architecture
- LightOnOCRForConditionalGeneration
- Context Length
- 16,384 tokens
- Vocabulary Size
- 151,936
- Release Date
- 2026-01-16
- License
- Apache 2.0
Frequently Asked Questions
How much VRAM does LightOnOCR 2 1B need?
LightOnOCR 2 1B requires 1.0 GB of VRAM at Q4_K_M, or 2.4 GB at BF16. Full 16K context adds up to 0.8 GB (1.8 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 1.0B × 4.8 bits ÷ 8 = 0.6 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 ≈ 1.2 GB (at full 16K context)
VRAM usage by quantization
What's the best quantization for LightOnOCR 2 1B?
For LightOnOCR 2 1B, Q4_K_M (1.0 GB) offers the best balance of quality and VRAM usage. Q5_K_S (1.1 GB) provides better quality if you have the VRAM. The smallest option is IQ2_XXS at 0.7 GB.
VRAM requirement by quantization
★ Recommended — best balance of quality and VRAM usage.
Can I run LightOnOCR 2 1B on a Mac?
Yes — MacBook Air 13" M3 (8 GB) and 145 other Macs can run LightOnOCR 2 1B. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 1.0 GB of usable unified memory (RAM minus macOS overhead).
Can I run LightOnOCR 2 1B locally?
Yes — LightOnOCR 2 1B can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.
How fast is LightOnOCR 2 1B?
At Q4_K_M, LightOnOCR 2 1B can reach ~4706 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~642 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 ÷ 1.02 × 0.65 = ~5098 tok/s
Estimated speed at Q4_K_M (1.0 GB)
Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
What's the download size of LightOnOCR 2 1B?
At Q4_K_M, the download is about 0.60 GB. The full-precision BF16 version is 2.01 GB. The smallest option (IQ2_XXS) is 0.28 GB.
Which GPUs can run LightOnOCR 2 1B?
52 consumer GPUs can run LightOnOCR 2 1B 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 LightOnOCR 2 1B?
163 devices with unified memory can run LightOnOCR 2 1B at Q4_K_M (1.0 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.