Instructions to use osunlp/QUEST-35B-MT-Plus-SFT with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use osunlp/QUEST-35B-MT-Plus-SFT with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="osunlp/QUEST-35B-MT-Plus-SFT")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# Load model directly
from transformers import AutoProcessor, AutoModelForImageTextToText

processor = AutoProcessor.from_pretrained("osunlp/QUEST-35B-MT-Plus-SFT")
model = AutoModelForImageTextToText.from_pretrained("osunlp/QUEST-35B-MT-Plus-SFT")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use osunlp/QUEST-35B-MT-Plus-SFT with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "osunlp/QUEST-35B-MT-Plus-SFT"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "osunlp/QUEST-35B-MT-Plus-SFT",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/osunlp/QUEST-35B-MT-Plus-SFT

SGLang

How to use osunlp/QUEST-35B-MT-Plus-SFT with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "osunlp/QUEST-35B-MT-Plus-SFT" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "osunlp/QUEST-35B-MT-Plus-SFT",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "osunlp/QUEST-35B-MT-Plus-SFT" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "osunlp/QUEST-35B-MT-Plus-SFT",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use osunlp/QUEST-35B-MT-Plus-SFT with Docker Model Runner:
```
docker model run hf.co/osunlp/QUEST-35B-MT-Plus-SFT
```

QUEST-35B-MT-Plus-SFT

File size: 3,019 Bytes

80a4c59
 
d5f54d7
80a4c59
 
 
 
 
 
 
 
8bd2a72
 
 
 
 
 
 
 
 
 
 
 
 
 
80a4c59
eba2ccc
 
 
 
 
 
 
a15c5e0
eba2ccc
 
 
 
 
 
80a4c59
 
 
 
 
 
 
 
8bd2a72
80a4c59
 
 
8bd2a72
d5f54d7
b5ce6ad
 
 
 
eba2ccc

---
library_name: transformers
license: apache-2.0
pipeline_tag: text-generation
tags:
- quest
- text-generation
---

# QUEST-35B-MT+SFT

QUEST **35B-class MoE** checkpoint after **mid-training + SFT** (Qwen3.5-35B-A3B base). Same training stage as QUEST-35B-MT; results correspond to the +MT column in the training-stage analysis.

## Benchmark results

| Benchmark | Metric | Score |
| --- | --- | ---: |
| BrowseComp | avg@3 | 45.5 |
| Mind2Web 2 | avg@3 | 29.9 |
| HLE | avg@3 | 39.74 |
| DeepResearch Bench | avg@3 | 39.72 |
| BrowseComp-Plus | avg@3 | 58.6 |
| WideSearch | Item F1 avg@4 | 62.5 |
| GAIA | avg@3 | 83.17 |
| LiveResearchBench | avg@3 | 65.47 |

## QUEST Family

| Type | Resources |
| --- | --- |
| 35B checkpoints | [RL](https://huggingface.co/osunlp/QUEST-35B-RL), [MT+SFT](https://huggingface.co/osunlp/QUEST-35B-MT-Plus-SFT), [MT](https://huggingface.co/osunlp/QUEST-35B-MT), [SFT](https://huggingface.co/osunlp/QUEST-35B-SFT) |
| 30B checkpoints | [RL](https://huggingface.co/osunlp/QUEST-30B-RL), [MT+SFT](https://huggingface.co/osunlp/QUEST-30B-MT-Plus-SFT), [SFT](https://huggingface.co/osunlp/QUEST-30B-SFT) |
| Smaller checkpoints | [9B](https://huggingface.co/osunlp/QUEST-9B), [4B](https://huggingface.co/osunlp/QUEST-4B), [2B](https://huggingface.co/osunlp/QUEST-2B) |
| Training data | [RL data](https://huggingface.co/datasets/osunlp/QUEST-RL-Data), [SFT objective data](https://huggingface.co/datasets/osunlp/QUEST-SFT-Data-Objective), [SFT open-ended data](https://huggingface.co/datasets/osunlp/QUEST-SFT-Data-Open-ended), [Mid-training data](https://huggingface.co/datasets/osunlp/QUEST-Mid-Training-Data) |

Model selection note: if you only need to evaluate objective tasks and do not
need open-ended task evaluation, we recommend the MT+SFT checkpoints because
they perform better on reasoning-heavy objective benchmarks. For a more comprehensive evaluation
across both objective and open-ended tasks, we recommend the RL checkpoints.

## Quick start

```python
from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "osunlp/QUEST-35B-MT-Plus-SFT"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(
    model_id, device_map="auto", torch_dtype="auto",
)
```

Apply the model's chat template with `tokenizer.apply_chat_template(...)` before passing prompts.

## License

Released under the **Apache License 2.0**.

## Citation

If our paper or related resources prove valuable to your research, we kindly ask
for a citation.

```bibtex
@misc{xie2026quest,
  title={QUEST: Training Frontier Deep Research Agents with Fully Synthetic Tasks},
  author={Xie, Jian and Lin, Tianhe and Wang, Zilu and Ning, Yuting and Yao, Yuekun and Xue, Tianci and Zhang, Zhehao and Li, Zhongyang and Zhang, Kai and Wu, Yufan and Chen, Shijie and Gou, Boyu and Han, Mingzhe and Wang, Yifei and Lee, Vint and Wei, Xinpeng and Wang, Xiangjun and Su, Yu and Sun, Huan},
  journal={arXiv preprint arXiv:2605.24218},
  year={2026}
}
```