CatalogWantedMirrorsLedgerOperating StatementDocsStudio statusCard system
alt.freeweights.minicpm / MiniCPM-OCR-1.0-DSpark-gguf
From: openbmb <huggingface.co/openbmb>Newsgroups: alt.freeweights.minicpmSubject:

MiniCPM-OCR-1.0-DSpark-gguf

Date: 2026-11-14Commit: cab493aSigned: 2DC28CC7B26A283B ✓

Image text to text model. Gguf weights, 2.98 GB in all. Published by openbmb under Apache-2.0.

minicpm2.98GBMiniCPM-OCR-1.0-DSpark-ggufggufpermissive
FactsParamsnot statedSize2.98 GBFormatsggufTaskimage text to textLicenseApache-2.0From sourcepermissive
$ fw get openbmb/MiniCPM-OCR-1.0-DSpark-gguf
# manifest sha256 003544addae2… · 2.98 GB · 15 files
$ fw verify ./MiniCPM-OCR-1.0-DSpark-gguf openbmb/MiniCPM-OCR-1.0-DSpark-gguf
From sourceHosted (signed)openbmb/MiniCPM-OCR-1.0-DSpark-ggufLast check: 2026-11-14 21:24 UTC
Run at homeGet help

Get help: custom fine-tune or private deploy. We review your setup before we quote.

Fine-tune of MiniCPM-OCR-1.0.

Works with

Run13Convert1Evaluate13Scan12Apps62

Convert with

ggml
OverviewModel cardFilesUse itLicenseProvenanceModel tree

Overview

Size
2.98 GB
Weights
209 MB (smallest of 11 GGUF variants)
Parameters
not stated
Formats
gguf
Task
image text to text
License
Apache-2.0 (permissive)
Files
15, each with a sha256
Publisher
openbmb
Added
2026-11-14
Section
alt.freeweights.minicpm
Lineage

Original release from the publisher.

Base model: openbmb/MiniCPM-OCR-1.0

Where these bytes come from
Publisher at the pinned commit→You

Sources are tried in order. A file counts only when its size and sha256 match the manifest. Sources are checked when you ask for them, not on a timer.

Sources
BadgeHostLast checkResult
From sourcehuggingface.co2026-11-14 21:24 UTCpassed

Model card

From the publisher at commit cab493a. Shown as plain text with links; images open at the publisher.

MiniCPM logo

Usage Guide

| πŸš€ Free API

Try it online for free, then call the API from your own code.

Playground (CN / Global)

API docs (CN / Global)

| πŸ€– Agent Skill

One prompt: the agent installs it and parses your first document.

SKILL.md

| πŸ’» Self-hosting

vLLM / SGLang / llama.cpp / Ollama, plus an official Docker image. Data stays on your side.

Deployment guide

| πŸ“± Mobile App

On-device and offline. Document parsing is iOS-only for now.

iOS Demo

Demo (CN / Global) &nbsp;Β·&nbsp; GitHub &nbsp;Β·&nbsp; SDK usage &nbsp;Β·&nbsp; DSpark &nbsp;Β·&nbsp; Discord &nbsp;|&nbsp; δΈ­ζ–‡

MiniCPM-OCR 1.0 DSpark β€” GGUF#

GGUF build of the DSpark draft head for MiniCPM-OCR 1.0. A draft model, not a standalone OCR model β€” pair it with the matching main checkpoint.

Files#

FileSize
dspark-MiniCPM-OCR-1_0-F16.gguf0.606 GB
dspark-MiniCPM-OCR-1_0-Q8_0.gguf0.327 GB
dspark-MiniCPM-OCR-1_0-Q6_K.gguf0.255 GB
dspark-MiniCPM-OCR-1_0-Q5_K_M.gguf0.233 GB
dspark-MiniCPM-OCR-1_0-Q5_K_S.gguf0.232 GB
dspark-MiniCPM-OCR-1_0-Q5_1.gguf0.243 GB
dspark-MiniCPM-OCR-1_0-Q5_0.gguf0.232 GB
dspark-MiniCPM-OCR-1_0-Q4_K_M.gguf0.211 GB
dspark-MiniCPM-OCR-1_0-Q4_K_S.gguf0.209 GB
dspark-MiniCPM-OCR-1_0-Q4_1.gguf0.220 GB
dspark-MiniCPM-OCR-1_0-Q4_0.gguf0.209 GB

Use the main model's projector (mmproj-model-f16-4x.gguf from MiniCPM-OCR-1.0-gguf); this repository contains only the draft.

Parsing pipeline#

MiniCPM-OCR 1.0 parses full documents as a two-stage pipeline orchestrated by the minicpm-ocr SDK. The two stages use separate models, run in separate environments, and may run on different machines:

  • Stage 1 β€” Layout detection (SDK environment). The PP-DocLayoutV3 layout model detects document regions β€” text, tables, formulas, charts, seals β€” and determines reading order; the SDK crops each region. It runs locally inside the SDK environment and requires PaddlePaddle / PaddleX.
  • Stage 2 β€” Region recognition (inference service). The SDK sends each cropped region over HTTP to a MiniCPM-OCR 1.0 inference service β€” served by vLLM, SGLang, llama.cpp, or Ollama β€” with a prompt matching the region type, then assembles the recognized content into the final Markdown / JSON in reading order.

Image: Parsing pipeline: in the minicpm-ocr SDK environment, the PP-DocLayoutV3 layout model detects and crops document regions; each cropped region is sent over HTTP to a separate recognition service running MiniCPM-OCR 1.0 (vLLM, SGLang, llama.cpp, or Ollama); the SDK then assembles the recognized text, tables, formulas, charts, and seals into Markdown and JSON in reading order.

See the model guide for more on architecture, parsing workflow, deployment options, and benchmark breakdowns.

Usage#

llama-cli \
  -m MiniCPM-OCR-1_0-Q8_0.gguf \
  --mmproj mmproj-model-f16-4x.gguf \
  -md dspark-MiniCPM-OCR-1_0-Q8_0.gguf \
  --spec-type draft-dspark \
  --image page.png -p "Text Recognition:" \
  --jinja -ngl 99 -ngld 99 -fa on

-md is the draft model and --spec-type draft-dspark enables the draft; -ngld offloads draft layers to GPU.

llama-mtmd-cli does not accept a draft model. Use llama-cli or llama-server when enabling DSpark with multimodal input.

Each tier carries the absolute target token IDs in its d2t tensor (the safetensors export stores them delta-encoded).

FormatHugging FaceModelScope
BF16 (original)πŸ€—ModelScope
GGUF (llama.cpp)πŸ€—ModelScope
AWQ W4A16πŸ€—ModelScope
GPTQ W4A16πŸ€—ModelScope
DSpark draft (BF16)πŸ€—ModelScope
DSpark draft (GGUF) (this repo)πŸ€—ModelScope

FAQ#

Installation, model downloads, GPU memory, input formats, and unexpected outputs: FAQ.

License#

Model weights, code, and documentation are released under the Apache License 2.0.

Files

15 files Β· 2.98 GBWeights 209 MB (smallest of 11 GGUF variants). Sizes and sha256 from the signed manifest.
Hide all 15 files with their sha256
FileSizesha256
.gitattributes2.29 KB90cc1cff20dc52fddeb11638e6eddb26fd70982377e5de33561597ca0380a163
README.md7.33 KBb61b12d8293e919fdaf560f37fcaa9be5417c20ee917a7a1f28658240b0a20d3
README_zh.md6.94 KBb8af3d6ad27e496eb931b4553cabbe6efbfd52b8664045513faa3a2310276e68
assets/parsing_pipeline.svg4.49 KBe601a364eeb37709a7d100e86b0099e57e7a071610bcaf22884d21fb1e0f127e
dspark-MiniCPM-OCR-1_0-F16.gguf606 MB293c40eee003e6d1a7bf502876b524d358c7d4dbb4b6266b1d892b9725f81f3d
dspark-MiniCPM-OCR-1_0-Q4_0.gguf209 MB7afa986d15c399fbee79c971de3e4922b24c0e50ccac5c24286eaa34be04bbfb
dspark-MiniCPM-OCR-1_0-Q4_1.gguf220 MBff02c46e95ca9222588b25d832203c74d153d0284fded54a212dc997f6f59022
dspark-MiniCPM-OCR-1_0-Q4_K_M.gguf211 MB34b8eeb83cff5fe920d89f21b34a207f02334e46ac73a41f07fa0f865b904239
dspark-MiniCPM-OCR-1_0-Q4_K_S.gguf209 MBd2ee7e1cc6e9825275c3de6f7a81f57c4fe818a12d6fa3789b44184f89489afd
dspark-MiniCPM-OCR-1_0-Q5_0.gguf232 MB591b5a4dd267f68c5a15e1206dc8eb6ebeaa4fc025f22dc294615543f40daef7
dspark-MiniCPM-OCR-1_0-Q5_1.gguf243 MB0999327d2b8fd240915a9b1c1db330c91f0b5c89118418cdec8a3140cb046a5f
dspark-MiniCPM-OCR-1_0-Q5_K_M.gguf233 MB6b0ff27d34655a3dfc29c4d8725a6a169c863682040ad7d3f2ff1500a13a0699
dspark-MiniCPM-OCR-1_0-Q5_K_S.gguf232 MB9023b393844e08d53bdc084106f6f2dd1c8e6f8e4812512619ead50f3450b873
dspark-MiniCPM-OCR-1_0-Q6_K.gguf255 MBd42f8f5cab648a22cc5db84256c7d86c6d8d9b055bc8c8338ba3b1107c8770ff
dspark-MiniCPM-OCR-1_0-Q8_0.gguf327 MBfd8f64bb62cf5a5c6b4c0999bfa58f64453690c9f75f1124ab613dfb44a822da
VersionsEach time the pinned upstream commit or file list changed in this repo. Newest first.
Upstream commitDateContents
cab493a5a2 pinned now manifest updated, same commit2026-11-14 21:24 UTC15 files Β· 2.98 GB
cab493a5a22026-11-14 21:24 UTC15 files Β· 2.98 GB

Use it

$ fw get openbmb/MiniCPM-OCR-1.0-DSpark-gguf
# manifest sha256 003544addae2… · 2.98 GB · 15 files
$ fw verify ./MiniCPM-OCR-1.0-DSpark-gguf openbmb/MiniCPM-OCR-1.0-DSpark-gguf
fw.pyChecks the signature and every sha256, tries each source in order, resumes.
$ python3 fw.py get https://freeweights.allternit.com/manifests/openbmb/MiniCPM-OCR-1.0-DSpark-gguf.json --dir ./MiniCPM-OCR-1.0-DSpark-gguf $ python3 fw.py verify ./MiniCPM-OCR-1.0-DSpark-gguf ./openbmb__MiniCPM-OCR-1.0-DSpark-gguf.json
hfThe Hugging Face CLI with one setting. Same repo id.
$ HF_ENDPOINT=https://freeweights.allternit.com hf download openbmb/MiniCPM-OCR-1.0-DSpark-gguf --revision cab493a5a26516a1c4c6e07c73b4470ff854a898 --local-dir ./MiniCPM-OCR-1.0-DSpark-gguf
llama.cppDownload one GGUF variant (dspark-MiniCPM-OCR-1_0-Q4_0, 209 MB), then serve it.
$ HF_ENDPOINT=https://freeweights.allternit.com hf download openbmb/MiniCPM-OCR-1.0-DSpark-gguf --include "dspark-MiniCPM-OCR-1_0-Q4_0*" --local-dir ./MiniCPM-OCR-1.0-DSpark-gguf $ llama-server -m ./MiniCPM-OCR-1.0-DSpark-gguf/dspark-MiniCPM-OCR-1_0-Q4_0.gguf
OllamaMake an Ollama model from the downloaded file.
$ printf 'FROM ./MiniCPM-OCR-1.0-DSpark-gguf/dspark-MiniCPM-OCR-1_0-Q4_0.gguf\n' > Modelfile $ ollama create minicpm-ocr-1.0-dspark-gguf -f Modelfile
Run at home

License

Name
Apache License 2.0
Id
Apache-2.0
Class
Permissive
Redistribution
yes
Listing
Hosted (signed)
Decided
License as declared by the publisher. No person read it. 2026-11-14
Text in
README.md
Text sha256
b61b12d8293e919fdaf560f37fcaa9be5417c20ee917a7a1f28658240b0a20d3
Terms gate
No extra terms beyond the license.

Read the license text

Licenses are checked against the license table. A custom license passes only after a person reads it.

Provenance

From: openbmb <huggingface.co/openbmb>Newsgroups: alt.freeweights.minicpmSubject: MiniCPM-OCR-1.0-DSpark-ggufDate: 2026-11-14Commit: cab493aSigned: 2DC28CC7B26A283B ✓
Upstream
openbmb/MiniCPM-OCR-1.0-DSpark-gguf
Pinned commit
cab493a5a26516a1c4c6e07c73b4470ff854a898
Manifest
MiniCPM-OCR-1.0-DSpark-gguf.json Β· .minisig
Manifest sha256
003544addae262b84feb3b0716e6b5643cdcddfd17d000f30d8a264e99793558
Signature
minisign, Ed25519, key 2DC28CC7B26A283B
Created
2026-11-14 21:24 UTC
Generator
freeweights-manifest/0.1.0
Takedowns
None.

How to verify

$ curl -O https://freeweights.allternit.com/manifests/openbmb/MiniCPM-OCR-1.0-DSpark-gguf.json $ curl -O https://freeweights.allternit.com/manifests/openbmb/MiniCPM-OCR-1.0-DSpark-gguf.json.minisig $ minisign -Vm MiniCPM-OCR-1.0-DSpark-gguf.json -P RWQ7KGqyx4zCLXtUFOqvfv8L3g+8fpKJ1FsuxPyOYMF+4KufbUxKq1Oq $ shasum -a 256 <file> # compare with the Files tab

A manifest whose signature fails must be rejected. A file whose size or sha256 does not match must be discarded. Verify a download

Model tree

Parent

openbmb/MiniCPM-OCR-1.0 Β· this model is a finetune of it

No model in the catalog names this one as its parent.

Get it
Download
$ HF_ENDPOINT=https://freeweights.allternit.com hf download openbmb/MiniCPM-OCR-1.0-DSpark-gguf
Open in Freeweights Studio

Pull, verify and run it in the Allternit app on Desktop, web or phone.

Signed manifest
minisign, key 2DC28CC7B26A283B. MiniCPM-OCR-1.0-DSpark-gguf.json. Provenance. Takedown history: none.
Source
huggingface.co/openbmb/MiniCPM-OCR-1.0-DSpark-gguf

Related models

All of alt.freeweights.minicpm β†’
From: openbmb <huggingface.co/openbmb>Subject: AgentCPM-Explore-GGUF
From: openbmb <huggingface.co/openbmb>Subject: AgentCPM-Report
From: openbmb <huggingface.co/openbmb>Subject: AgentCPM-Report-GGUF
From: openbmb <huggingface.co/openbmb>Subject: BitCPM-CANN-0.5B-gguf
From: openbmb <huggingface.co/openbmb>Subject: BitCPM-CANN-1B-gguf
From: openbmb <huggingface.co/openbmb>Subject: BitCPM-CANN-3B-gguf