---
license: apache-2.0
library_name: pytorch
pipeline_tag: image-classification
tags:
  - decision-model
  - vision
  - multimodal
  - system-one
  - s1mb
  - darwin
  - ztc
  - custom-code
---

# Darwin-35B-ZTC-Vision

<p align="center">
  <a href="https://huggingface.co/spaces/hotchpotch/S1MB-leaderboard?category=vision-v1"><img src="https://img.shields.io/badge/S1MB_vision--v1-merged_2026--10--19-gold?style=for-the-badge" alt="S1MB vision-v1"></a>
  <a href="https://huggingface.co/FINAL-Bench/Darwin-27B-ZTC-v9"><img src="https://img.shields.io/badge/🧬_Text-Darwin--27B--ZTC--v9-blue?style=for-the-badge" alt="Text model"></a>
</p>
<p align="center">
  <a href="https://huggingface.co/FINAL-Bench/Darwin-35B-A3B-Opus"><img src="https://img.shields.io/badge/🧬_Backbone-Darwin--35B--A3B--Opus-blue?style=for-the-badge" alt="Backbone"></a>
  <a href="https://huggingface.co/FINAL-Bench/Darwin-27B-ZTC-v5"><img src="https://img.shields.io/badge/🧬_Series-Darwin--27B--ZTC--v5-blue?style=for-the-badge" alt="ZTC series"></a>
  <a href="https://huggingface.co/FINAL-Bench"><img src="https://img.shields.io/badge/🏠_Org-FINAL--Bench-black?style=for-the-badge" alt="FINAL-Bench"></a>
  <a href="https://www.apache.org/licenses/LICENSE-2.0"><img src="https://img.shields.io/badge/License-Apache_2.0-green?style=for-the-badge" alt="License"></a>
</p>

Darwin-35B-ZTC-Vision is a vision decision model. It does not generate text. Given one or more images, a state and a typed question, it returns a calibrated probability for every option in one forward pass: yes or no (`noul`), one of several options (`choice`), or a level on an ordered scale (`score`). It handles up to four images per request, for example "which of the two photos has the higher quality" or "is there a defect on this part".

It is a 35B mixture-of-experts model (about 3B active parameters per token) built on [Darwin-35B-A3B-Opus](https://huggingface.co/FINAL-Bench/Darwin-35B-A3B-Opus) with the Darwin ZTC decision readout. It is the vision member of the ZTC series; the text member is [Darwin-27B-ZTC-v9](https://huggingface.co/FINAL-Bench/Darwin-27B-ZTC-v9).

## Results

### S1MB (System One Mosaic Benchmark), vision-v1

All 20 required vision benchmarks complete (the optional Box tasks are not supported), zero failed decisions. Merged by the maintainer on 2026-10-19 ([PR #76](https://huggingface.co/datasets/hotchpotch/s1mb-result/discussions/76)). At submission it ranked first (Borda 91.00) on the published vision results of 2026-10-17 (16 complete models) with this model added, recomputed with the leaderboard's own Borda method; that table is below. More models have entered since: on the published results of 2026-10-19 (32 complete models) the same recomputation places it fifth with 81.80. The [live vision leaderboard](https://huggingface.co/spaces/hotchpotch/S1MB-leaderboard?category=vision-v1) is authoritative.

| Rank (2026-10-17 data) | Model | Borda score | Mean baseline-adjusted score |
|---|---|---|---|
| **1** | **Darwin-35B-ZTC-Vision** | **91.00** | **73.98** |
| 2 | Jev 1.14 (Preview) | 90.00 | 74.64 |
| 3 | Perplexity Decider Flash 1.2 | 77.50 | 67.35 |
| 4 | Imajev 4B | 67.83 | 69.29 |
| 5 | Clef-Flash | 63.50 | 63.70 |

### Selected benchmarks (baseline-adjusted score)

| Benchmark | Darwin-35B-ZTC-Vision | Jev 1.14 (Preview) |
|---|---|---|
| tallyqa-score (counting) | 0.720 | 0.546 |
| visa-noul (anomaly detection) | 0.603 | 0.467 |
| koniq-score (image quality) | 0.196 | 0.143 |
| cvbench-relation-choice | 0.965 | 0.944 |
| pope-noul (object hallucination) | 0.793 | 0.767 |
| nights-choice (perceptual similarity, 3 images) | 0.886 | 0.866 |
| vsr-noul (spatial relations) | 0.775 | 0.768 |
| generalization-vision-easy-choice | 0.976 | 0.967 |

## Training and data

Full-weight fine-tuning with a decision readout on public **train** splits and our own synthetic data:

- COCO train2017 images (licensed subset) for verification, counting, relation and multi-hop spatial questions computed from the annotations, plus synthetic quality degradations rated on a five-level scale.
- VisA normal images from the official train split (CC BY 4.0) with synthetic surface defects, plus procedural textures.
- Public train splits of text and chart decision datasets, and the text decision data of the Darwin ZTC series.

No S1MB question text was used for training.

## Usage

Python 3.12, a CUDA GPU with about 75 GB of free memory (BF16 weights), and the code shipped in this repository (`autojev/`).

```python
import sys
from pathlib import Path
from huggingface_hub import snapshot_download

path = Path(snapshot_download("FINAL-Bench/Darwin-35B-ZTC-Vision"))
sys.path.insert(0, str(path))
from autojev.model import DecisionModel, answer

model = DecisionModel(checkpoint=path).eval()
question = {
    "type": "noul",
    "instructions": "Determine whether image_1 shows a defect such as a scratch, crack, hole or stain.",
    "criteria": {"true": "A defect is present", "false": "The surface is normal"},
}
row = {"state": {"images": ["image_1"]}, "images": ["part.jpg"], "question": question}
print(answer(question, model.predict([row])[0]))
```

`decide_server.py` serves `POST /v1/systemone` (images as data URLs, up to four per request):

```bash
ZTC_MODEL=/path/to/snapshot PORT=8410 python decide_server.py
```

`s1mb_adapter/ztc_http.py` is the S1MB evaluator adapter we used to measure the results above against this server.

## License

Apache-2.0 for the weights and code in this repository. Training images keep their own licenses (COCO: per-image Flickr licenses and CC BY 4.0 annotations; VisA: CC BY 4.0).