# KORMo-Web-10B
**KORMo-Web-10B** is a Korean-first web agent built entirely on the **KORMo** model family, which KAIST MLP Lab trains **from scratch**. It reads a screenshot, thinks, and takes one browser action at a time (click, type, scroll, move between tabs and sites) until the task is done.
Among the open web agents we compared, it scores the **highest in Korean (30.1%)** on Online-Mind2Web, and **29.0% in English**.
## From a Korean LLM to a Korean web agent
KORMo-Web-10B is the third step of one Korean lineage, not a fine-tune of a foreign base model:
| Step | Model | What it adds |
|---|---|---|
| 1 | [KORMo-10B](https://huggingface.co/KORMo-Team/KORMo-10B-sft) | Korean–English LLM trained from scratch, with its own Korean tokenizer |
| 2 | [KORMo-VL](https://huggingface.co/KORMo-VL/KORMo-VL) | Vision: SigLIP encoder + KORMo-10B, trained on Korean document and UI images |
| 3 | **KORMo-Web-10B** | Web agent: large-scale web understanding data and browser trajectories, in Korean and English |
Because every stage was trained with Korean as a first language, the model handles Korean pages, Korean UI terms and natural Korean requests directly.
## Performance
All models below ran **the same 612 live-website tasks** (300 English, 312 Korean), each with its own prompt and action format, and were scored by **the same LLM judge**.
| Model | Size | Online-Mind2Web EN | Online-Mind2Web KO |
|---|---:|---:|---:|
| Qwen3.5-9B (base, no fine-tuning) | 9B | 24.7 | 16.7 |
| Fara-1.5-4B | 4B | 20.7 | 7.7 |
| UI-TARS-1.5-7B | 7B | 22.3 | 20.2 |
| OpenWebRL-4B | 4B | 32.7 | 15.4 |
| **KORMo-Web-10B (ours)** | **11B** | **29.0** | **30.1** |
| GPT-6-luna (proprietary) | – | 63.0 | 65.7 |
| Gemini-3.8-flash (proprietary) | – | 74.0 | 80.1 |
Success rate (%). The denominator is always the full task count, so crashed or empty runs count as failures.
KORMo-Web-10B has the best Korean score among the open web agents above, by about 10 points.
## Why KORMo-Web-10B
- **Korean from the ground up.** Built on a Korean LLM trained from scratch, not adapted from an English-first model.
- **Korean and English, natural requests.** Handles short commands as well as long, conversational instructions in either language.
- **Screenshot-only.** It needs no DOM or accessibility tree; it acts on what is on screen and checks each action's result on the next screenshot.
- **Thinks, then acts.** A short reasoning step precedes every action, so its behaviour is easy to inspect.
- **Open and ready to run.** Weights, model code for transformers 5, a vLLM path, a reference browser harness and the exact inference settings are all included.
## Quickstart
Tested with **transformers 5.12.1** and **vLLM 0.23.0**. The model runs on a single GPU.
transformers does not ship the KORMo text model, so this repo includes it (`_modeling_kormo.py`) together with a small loader, `kormo_load.py`:
- `kormo_load.register(path)` registers the KORMo text model. Call it **before** loading the processor or the model.
- `kormo_load.patch_vision(model)` feeds the SigLIP features *after* post-layernorm to the projector, as in training. Without it the model still runs but misreads screenshots.
### Inference example (one step, transformers)
```bash
pip install -U "transformers>=5.12" torch pillow huggingface_hub
huggingface-cli download KORMo-VL/KORMo-Web-10B --local-dir KORMo-Web-10B
```
```python
import json, sys, torch
from PIL import Image
from transformers import AutoModelForImageTextToText, AutoProcessor
M = "KORMo-Web-10B"
sys.path.insert(0, M)
import kormo_load
kormo_load.register(M)
processor = AutoProcessor.from_pretrained(M)
model = AutoModelForImageTextToText.from_pretrained(M, dtype=torch.bfloat16).to("cuda").eval()
kormo_load.patch_vision(model)
SYSTEM = open(f"{M}/system_prompt.md").read().strip()
TOOLS = json.load(open(f"{M}/tools.json"))
obs = ("
\nscreen size: 1280 x 720\nCurrent tab: 0\nAvailable tabs:\n"
"- Tab 0 (active): https://www.aladin.co.kr/\n\nurl: https://www.aladin.co.kr/\nscreenshot:\n")
messages = [
{"role": "system", "content": SYSTEM},
{"role": "user", "content": [
{"type": "text", "text": "\n알라딘 사이트에서 '해커스 토익 RC'를 검색한 뒤 가장 먼저 표시되는 도서의 출간일을 알려줘.\n\n" + obs},
{"type": "image", "image": Image.open("screen.png").convert("RGB")},
{"type": "text", "text": "\n"},
]},
]
inputs = processor.apply_chat_template(messages, tools=TOOLS, add_generation_prompt=True,
tokenize=True, return_dict=True, return_tensors="pt").to(model.device)
out = model.generate(**inputs, max_new_tokens=16384, do_sample=True, temperature=0.6, top_p=0.95, top_k=20)
print(processor.decode(out[0][inputs["input_ids"].shape[1]:], skip_special_tokens=False))
```
A real first step from our Online-Mind2Web run (Aladin home page):
```text
I see the Aladin homepage, with the search field labeled “통합검색” near the top. I have not yet searched for the book, so I still need to enter “해커스 토익 RC,” inspect the first result’s publication date, and report that date. My plan:
1. Click the search field
2. Enter “해커스 토익 RC” and submit the search
3. Read the publication date of the first listed book
This click targets the visible search field and is a safe, relevant first action.
482
212
```
Coordinates are normalized to 0–1000. Convert them to pixels with `x / 1000 × width` and `y / 1000 × height`.
### Run the agent (vLLM)
vLLM does not include the KORMo architecture either. The KORMo text layers have the Llama structure, so `vllm/convert_to_vllm.py` renames two parameter names and writes a folder vLLM loads natively. `vllm/sitecustomize.py` turns SigLIP post-layernorm back on inside vLLM.
**1. Convert (once)**
```bash
pip install -U "vllm>=0.23.0" openai playwright safetensors
playwright install chromium
python KORMo-Web-10B/vllm/convert_to_vllm.py KORMo-Web-10B KORMo-Web-10B-vllm
```
**2. Start the model server**
```bash
PYTHONPATH=$PWD/KORMo-Web-10B/vllm vllm serve ./KORMo-Web-10B-vllm --served-model-name KORMo-Web-10B \
--max-model-len 65536 --limit-mm-per-prompt '{"image":1,"video":0}'
```
Check that the server log prints `[kormo-vllm] SigLIP post_layernorm enabled`.
**3. Run a browser task**
```bash
python KORMo-Web-10B/examples/agent_playwright.py --model KORMo-Web-10B \
--task "다나와에서 50만 원 이하, 10인치 이상 태블릿 3개를 찾아서 가격을 알려줘"
```
The agent opens Chromium, prints each action as it goes, and finishes with the model's answer.
## How It Works
```
user
task + observation (tabs, url, screenshot)
assistant
what I see · what is done · plan one action
tool action result + new observation (newest screenshot only)
… repeat until finished(answer)
```
- **Action space:** 22 browser tools: click, type, hotkey, scroll, drag, goto, tab control, `finished`, and others ([`tools.json`](tools.json)).
- **Memory:** the full text history of thoughts, actions and observations is kept. Only the current screenshot is sent.
- **Recommended settings:** temperature 0.6, top_p 0.95, top_k 20, thinking on, up to 16k tokens per step and 50 steps. These are already set in `generation_config.json`.
The exact message format, parsing rules and feedback strings are in **[HARNESS.md](HARNESS.md)**. Follow it if you build your own harness.
## Evaluation Details
| | |
|---|---|
| Tasks | Online-Mind2Web: 300 English, 312 Korean |
| Judge | GPT-6 |
## Files
| File | Description |
|---|---|
| `model.safetensors` · `config.json` | Weights and config (LLaVA-OneVision: SigLIP + KORMo-10B text model) |
| `_configuration_kormo.py` · `_modeling_kormo.py` | KORMo text model code for transformers 5 |
| `kormo_load.py` | `register()` and `patch_vision()` |
| `chat_template.jinja` | Chat template (XML tool calls, thinking on) |
| `system_prompt.md` · `tools.json` | System prompt and 22-tool schema |
| `HARNESS.md` | Full harness specification |
| `examples/agent_playwright.py` | Reference browser agent (Playwright + OpenAI-compatible API). It does not execute `upload`. |
| `vllm/convert_to_vllm.py` · `vllm/sitecustomize.py` | Convert for vLLM and the post-layernorm fix |
| `generation_config.json` · `*processor_config.json` · `tokenizer*.json` · `special_tokens_map.json` | Sampling, processor and tokenizer configs |
## 한국어 요약
**KORMo-Web-10B**는 KAIST MLP 연구실이 처음부터 학습한 한국어 모델 **KORMo-10B → KORMo-VL**을 이어 만든 한국어 우선 웹 에이전트입니다. 스크린샷만 보고 한국어·영어 웹 작업을 수행합니다.
- **Online-Mind2Web:** EN 29.0, KO 30.1
- 비교한 공개 웹 에이전트 중 한국어 점수가 가장 높습니다.
## Acknowledgements
Built on **[KORMo-VL](https://huggingface.co/KORMo-VL/KORMo-VL)** and **[KORMo-10B](https://huggingface.co/KORMo-Team/KORMo-10B-sft)** by KAIST MLP Lab. Evaluated on **Online-Mind2Web** and its Korean counterpart.