PA

saikei

PaddleocrGGUF by DaniReyUpdated 2026-10-30
Download on Hugging Face
Parameters34.7B
Context32K
Layers28
Architectureqwen3
Smallest quant0.6 GB
Quick answer

saikei needs about 2.8 GB of memory at its main file (a 1.3 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 5 quants, from 0.6 GB (Q8_0) to 12.1 GB. Longer context needs more: at 32K tokens the KV cache alone adds about 3.8 GB.

Saikei – a document‑reading assistant with vision and reasoning

Saikei is a 34.7B‑parameter Qwen3‑based model that can read, summarize, and process documents. It supports reasoning, tool use, and vision tasks. The model is released under the Apache‑2.0 license.

Good at

  • Document reading and summarization
  • Vision‑enabled document parsing
  • Reasoning and tool integration

Limits

  • No vision support on non‑supported platforms
  • Requires large VRAM for full context
  • License is Apache‑2.0

Best for

  • Summarizing PDFs and scanned documents
  • Extracting data from invoices
  • Automating document workflows

Who it’s for. Developers building desktop assistants that need to read and analyze documents locally. The model runs entirely on the user’s machine and does not send data to external servers.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q8_00.6 GB 2.1 GB
Fits Get file →
—0.9 GB 2.4 GB
Fits Get file →
—1.3 GB 2.8 GB
Fits Get file →
—10.1 GB 11.6 GB
Too big Get file →
—12.1 GB 13.5 GB
Too big Get file →

For 8 GB VRAM, use the Q8_0 quantized file (0.64 GB, ~2.1 GB memory at 8 k context). Larger VRAM allows the 0.94 GB or 1.33 GB files, which use ~2.4 GB or ~2.8 GB memory respectively. Estimates are approximate.

What GPU do you need for saikei?

These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.

  • 4 GB VRAM: the best quant that fits is the main file (1.3 GB file, about 2.8 GB in use).
  • 6 GB VRAM: the best quant that fits is the main file (1.3 GB file, about 2.8 GB in use).
  • 8 GB VRAM: the best quant that fits is the main file (1.3 GB file, about 2.8 GB in use).
  • 12 GB VRAM: the best quant that fits is the main file (10.1 GB file, about 11.6 GB in use).
  • 16 GB VRAM: the best quant that fits is the main file (12.1 GB file, about 13.5 GB in use).

Which saikei quant should you choose?

If memory is tight, the smallest file is Q8_0 at 0.6 GB, with the quality trade-offs that come with fewer bits. Close to lossless. It costs roughly twice the memory of a 4-bit file, so it suits cases where you have plenty of VRAM.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architectureqwen3
Parameters34.7B
Layers28
Trained context32,768 tokens
KV cache (fp16)112 KB per token

Source

Quantized byDaniRey
FamilyPaddleocr
RepositoryDaniRey/saikei
Licenseapache-2.0
Downloads (HF)1,120
Last updated2026-10-30

Frequently asked questions about saikei

How much VRAM does saikei need?

About 2.8 GB at its main file with an 8K context window. That is the 1.3 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.

Can I run saikei on an 8 GB GPU?

Yes, with the smaller files: the 1.3 GB file needs about 2.8 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of saikei GGUF?

It depends on the quant: Q8_0 is 0.6 GB, file is 0.9 GB, file is 1.3 GB, file is 10.1 GB, file is 12.1 GB.

What context length does saikei support?

The GGUF header lists a trained context length of 32,768 tokens (32K). The KV cache grows with the context you use: about 0.9 GB at 8K and 3.8 GB at 32K.

Where can I download saikei GGUF?

The files are on Hugging Face in the DaniRey/saikei repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

What license does saikei use?

Hugging Face lists the license as apache-2.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.

Can I run Saikei on 8 GB VRAM?

Yes, the Q8_0 quantized file (0.64 GB) uses about 2.1 GB memory at 8 k context, fitting within 8 GB VRAM.

Does Saikei support vision tasks?

Yes, Saikei includes vision capabilities for document parsing and image analysis.

Other GGUFs of PaddleOCR-VL-1.5

Similar models

Browse all →

More Paddleocr models

All Paddleocr →

Other large models

All large models →