Qianfan OCR — Hardware Requirements & GPU Compatibility
Vision
Qianfan-OCR is Baidu's 4.7-billion-parameter vision-language model for document intelligence rather than general chat. It pairs a Qianfan-ViT vision encoder with a Qwen3-4B language backbone, doing direct image-to-Markdown conversion alongside table extraction, chart understanding, and document Q&A in one end-to-end model instead of a multi-stage pipeline. It also has an optional "Layout-as-Thought" mode that reasons about page layout before producing output. At under 5 billion parameters, it runs on a single consumer or prosumer GPU once quantized. Context length is 32,768 tokens, extendable further per Baidu's documentation. It is released under the Apache 2.0 license, permitting unrestricted commercial and research use. Published in March 2026, Baidu reports it as its top-scoring end-to-end model on public document-parsing benchmarks, supporting 192 languages.
250.3K downloads
1.2K likesSpecifications
- Publisher
- Baidu
- Parameters
- 4.7B
- Architecture
- QianfanOCRForConditionalGeneration
- Context Length
- 32,768 tokens
- Vocabulary Size
- 153,678
- Release Date
- 2026-03-18
- License
- Apache 2.0
Get Started
HuggingFace
How Much VRAM Does Qianfan OCR Need?
Select a quantization to see compatible GPUs below.
| Quantization | Bits | VRAM | + Context | File Size | Quality |
|---|---|---|---|---|---|
| Q2_K | 3.40 | 2.5 GB | 5.3 GB | 2.02 GB | 2-bit quantization with K-quant improvements |
| Q3_K_S | 3.50 | 2.6 GB | 5.4 GB | 2.07 GB | 3-bit small quantization |
| Q3_K_M | 3.90 | 2.8 GB | 5.6 GB | 2.31 GB | 3-bit medium quantization |
| Q4_0 | 4.00 | 2.9 GB | 5.7 GB | 2.37 GB | 4-bit legacy quantization |
| Q4_K_M | 4.80 | 3.3 GB | 6.2 GB | 2.84 GB | 4-bit medium quantization — most popular sweet spot |
| Q5_K_M | 5.70 | 3.9 GB | 6.7 GB | 3.38 GB | 5-bit medium quantization — good quality/size tradeoff |
| Q6_K | 6.60 | 4.4 GB | 7.2 GB | 3.91 GB | 6-bit quantization, very good quality |
| Q8_0 | 8.00 | 5.2 GB | 8.1 GB | 4.74 GB | 8-bit quantization, near-lossless |
Which GPUs Can Run Qianfan OCR?
Q4_K_M · 3.3 GBQianfan OCR (Q4_K_M) requires 3.3 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 5+ GB is recommended. Using the full 33K context window can add up to 2.8 GB, bringing total usage to 6.2 GB. 52 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti.
Runs great
— Plenty of headroomWhich Devices Can Run Qianfan OCR?
Q4_K_M · 3.3 GB59 devices with unified memory can run Qianfan OCR, including NVIDIA DGX H100, NVIDIA DGX A100 640GB.
Runs great
— Plenty of headroomWhere to Download Qianfan OCR
Community quantizations of this model — GGUF for llama.cpp, Ollama, and LM Studio, plus AWQ/MLX variants where available.
Related Models
Frequently Asked Questions
- How much VRAM does Qianfan OCR need?
Qianfan OCR requires 3.3 GB of VRAM at Q4_K_M, or 10.0 GB at BF16. Full 33K context adds up to 2.8 GB (6.2 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 4.7B × 4.8 bits ÷ 8 = 2.8 GB
KV Cache + Overhead ≈ 0.5 GB (at 2K context + ~0.3 GB framework)
KV Cache + Overhead ≈ 3.4 GB (at full 33K context)
VRAM usage by quantization
Q4_K_M3.3 GBQ4_K_M + full context6.2 GB- What's the best quantization for Qianfan OCR?
For Qianfan OCR, Q4_K_M (3.3 GB) offers the best balance of quality and VRAM usage. Q5_K_S (3.8 GB) provides better quality if you have the VRAM. The smallest option is Q2_K at 2.5 GB.
VRAM requirement by quantization
Q2_K2.5 GBQ4_02.9 GBQ4_K_M ★3.3 GBQ5_K_S3.8 GBQ6_K4.4 GBBF1610.0 GB★ Recommended — best balance of quality and VRAM usage.
- Can I run Qianfan OCR on a Mac?
Qianfan OCR requires at least 2.5 GB at Q2_K, which exceeds the unified memory of most consumer Macs. You would need a Mac Studio or Mac Pro with a high-memory configuration.
- Can I run Qianfan OCR locally?
Yes — Qianfan OCR can run locally on consumer hardware. At Q4_K_M quantization it needs 3.3 GB of VRAM. Popular tools include Ollama, LM Studio, and llama.cpp.
- How fast is Qianfan OCR?
At Q4_K_M, Qianfan OCR can reach ~1441 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~197 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 ÷ 3.3 × 0.65 = ~1562 tok/s
Estimated speed at Q4_K_M (3.3 GB)
~1562 tok/s~197 tok/s~1562 tok/s~1441 tok/sReal-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
- What's the download size of Qianfan OCR?
At Q4_K_M, the download is about 2.84 GB. The full-precision BF16 version is 9.48 GB. The smallest option (Q2_K) is 2.02 GB.
- Which GPUs can run Qianfan OCR?
52 consumer GPUs can run Qianfan OCR at Q4_K_M (3.3 GB). 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 Qianfan OCR?
59 devices with unified memory can run Qianfan OCR at Q4_K_M (3.3 GB), including AMD Ryzen AI 9 HX 370 (Strix Point) Laptop, ASUS Ascent GX10, Apple iPhone 17 Pro, Asus ROG Flow Z13 (2025, 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.