Surya Ocr 2 — Hardware Requirements & GPU Compatibility
VisionSurya OCR 2 is Datalab's roughly 650-million-parameter document model, built on an architecture similar to Qwen3.5 for vision-language tasks. It manages OCR, layout analysis with reading order, and table recognition in one go, outputting either layout JSON or full-page HTML based on the prompt. The card reports a score of 83.3% on olmOCR-bench, noting this as the top result under 3 billion parameters, with performance around 5 pages per second on an RTX 5090. At this size, it runs easily on standard consumer GPUs. The context length reaches 262,144 tokens. The code uses the Apache 2.0 license, while weights follow a modified AI Pubs Open RAIL-M agreement described as free for research, personal projects, and startups with less than $5M in funding or revenue. Commercial use beyond that requires a separate deal from Datalab. It was released in May 2026.
Runs great on a 24 GB GPU
Runs from 8 GB (Q4_K_M · 256K)Best on 24 GB (Q4_K_M · ~862 tok/s · 256K)Macs from 8 GB (Q4_K_M · ~94 tok/s)
openrail
Reference GPU · RTX 4090 24 GB
Estimates, ±20% · how we estimate
Run Surya Ocr 2
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
vllm serve datalab-to/surya-ocr-2 --max-model-len 16384vLLM repo with its own quantisation
Download: Weights on Hugging Face
Ways to runOther ways
each opens step-by-stepRun in cloud
Runs on RTX 3060 12GB (11 GB headroom) · Q4_K_M
- Generation speed
- ~308 tok/s
- generation speed
- Cost per 1M output tokens
- $0.057
- per 1M output tokens
How Much VRAM Does Surya Ocr 2 Need?
Select a quantization level; the listed hardware adheres to it.
| Quantization | VRAM | +32K | On RTX 4090 (reference) | |
|---|---|---|---|---|
| 0.6 GB | 0.8 GB | 2-bit · K-quant — 2-bit quantization with K-quant improvements | Fits: 256K context · ~1024 tok/s | |
| 0.7 GB | 0.8 GB | 3-bit medium — 3-bit medium quantization | Fits: 256K context · ~964 tok/s | |
| 0.8 GB | 0.9 GB | 4-bit medium — 4-bit medium quantization — most popular sweet spot | Fits: 256K context · ~862 tok/s | |
| 0.8 GB | 1 GB | 5-bit medium — 5-bit medium quantization — good quality/size tradeoff | Fits: 256K context · ~780 tok/s | |
| 0.9 GB | 1.1 GB | 6-bit, very good — 6-bit quantization, very good quality | Fits: 256K context · ~712 tok/s | |
| 1.0 GB | 1.2 GB | 8-bit, near-lossless — 8-bit quantization, near-lossless | Fits: 256K context · ~630 tok/s | |
| 1.7 GB | 1.9 GB | 16-bit · bfloat — Brain floating point 16 — preferred for training | Fits: 256K context · ~381 tok/s |
Selected Q4_K_M
4-bit medium quantization0.8GBat 16K context
- Weights
- 0.4 GB
- KV cache
- 0.1 GB
- Overhead
- 0.3 GB
Which Hardware Can Run Surya Ocr 2?
Q4_K_M · 0.8 GBSurya Ocr 2 (Q4_K_M) requires 0.8 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 1+ GB is recommended. Using the full 262K context window can add up to 1.6 GB, bringing total usage to 2.3 GB. 114 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti. 185 devices with unified memory can run Surya Ocr 2, including Mac Studio M5 Ultra (256 GB), Mac Studio M5 Ultra (512 GB).
≤ 11 GB
≤ 11 GB tier →fits · 3.2 GB free · up to 256K ctx
68 GPUs · 1 Mac class · 3 device families
RTX 3080~650Simulate output on RTX 3080RTX 2080 Ti~527Simulate output on RTX 2080 TiRTX 3070 Ti~520Simulate output on RTX 3070 TiM3*8 GB~94Simulate output on iMac 24" M3 (8 GB)Orin Nano~87Simulate output on NVIDIA Jetson Orin Nano 8GB (Super)iPhone~63Simulate output on iPhone 1720 GB
fits · 19.2 GB free · up to 256K ctx
1 GPU
RX 7900 XT~632Simulate output on RX 7900 XT32–40 GB
fits · 31.2 GB free · up to 256K ctx
1 GPU · 3 Mac classes · 9 device families
RTX 5090~1533Simulate output on RTX 5090M5 Max*36 GB~424Simulate output on Mac Studio M5 Max (36 GB)M632 GB~157Simulate output on iMac 24" M6 (32 GB)RTX Spark N1X~257Simulate output on NVIDIA RTX Spark N1X Laptop (32 GB)Orin~175Simulate output on NVIDIA Jetson AGX Orin 32GB
By VRAM tier · typical GPUs and Macs at Q4_K_M
Specifications
- Publisher
- datalab-to
- Parameters
- 686M
- Architecture
- Qwen3_5ForConditionalGeneration
- Context Length
- 262,144 tokens
- Vocabulary Size
- 65,425
- Release Date
- 2026-05-14
- License
- openrail
Frequently Asked Questions
How much VRAM does Surya Ocr 2 need?
Surya Ocr 2 requires 0.8 GB of VRAM at Q4_K_M, or 1.7 GB at BF16. Full 262K context adds up to 1.6 GB (2.3 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 686M × 4.8 bits ÷ 8 = 0.4 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 ≈ 1.9 GB (at full 262K context)
VRAM usage by quantization
What's the best quantization for Surya Ocr 2?
For Surya Ocr 2, Q4_K_M (0.8 GB) offers the best balance of quality and VRAM usage. Q5_K_M (0.8 GB) provides better quality if you have the VRAM. The smallest option is Q2_K at 0.6 GB.
VRAM requirement by quantization
★ Recommended — best balance of quality and VRAM usage.
Can I run Surya Ocr 2 on a Mac?
Yes — MacBook Air 13" M3 (8 GB) and 158 other Macs can run Surya Ocr 2. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 0.8 GB of usable unified memory (RAM minus macOS overhead).
Can I run Surya Ocr 2 locally?
Yes — Surya Ocr 2 can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.
How fast is Surya Ocr 2?
At Q4_K_M, Surya Ocr 2 can reach ~6316 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~862 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 ÷ 0.76 × 0.65 = ~6842 tok/s
Estimated speed at Q4_K_M (0.8 GB)
Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
What's the download size of Surya Ocr 2?
At Q4_K_M, the download is about 0.41 GB. The full-precision BF16 version is 1.37 GB. The smallest option (Q2_K) is 0.29 GB.
Which GPUs can run Surya Ocr 2?
114 consumer GPUs can run Surya Ocr 2 at Q4_K_M. Top options include AMD Radeon RX 5600 XT, AMD Radeon RX 5700 XT, AMD Radeon RX 580 8GB. 114 GPUs have plenty of headroom for comfortable inference.
Which devices can run Surya Ocr 2?
185 devices with unified memory can run Surya Ocr 2 at Q4_K_M (0.8 GB), including AMD Ryzen 7 8845HS (Hawk Point) Laptop, AMD Ryzen AI 7 350 (Krackan Point) Laptop, AMD Ryzen AI 9 HX 370 (Strix Point) Laptop, ASUS Ascent GX10. Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.