OpenThinkerAgent-32B-SFT-100K

Instructions to use Artificial-Production-Units/OpenThinkerAgent-32B-SFT-100K with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Artificial-Production-Units/OpenThinkerAgent-32B-SFT-100K with Transformers:

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

pipe = pipeline("text-generation", model="Artificial-Production-Units/OpenThinkerAgent-32B-SFT-100K")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForMultimodalLM

tokenizer = AutoTokenizer.from_pretrained("Artificial-Production-Units/OpenThinkerAgent-32B-SFT-100K")
model = AutoModelForMultimodalLM.from_pretrained("Artificial-Production-Units/OpenThinkerAgent-32B-SFT-100K")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.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(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Artificial-Production-Units/OpenThinkerAgent-32B-SFT-100K with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Artificial-Production-Units/OpenThinkerAgent-32B-SFT-100K"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Artificial-Production-Units/OpenThinkerAgent-32B-SFT-100K",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Artificial-Production-Units/OpenThinkerAgent-32B-SFT-100K

SGLang

How to use Artificial-Production-Units/OpenThinkerAgent-32B-SFT-100K 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 "Artificial-Production-Units/OpenThinkerAgent-32B-SFT-100K" \
    --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": "Artificial-Production-Units/OpenThinkerAgent-32B-SFT-100K",
		"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 "Artificial-Production-Units/OpenThinkerAgent-32B-SFT-100K" \
        --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": "Artificial-Production-Units/OpenThinkerAgent-32B-SFT-100K",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Artificial-Production-Units/OpenThinkerAgent-32B-SFT-100K with Docker Model Runner:
```
docker model run hf.co/Artificial-Production-Units/OpenThinkerAgent-32B-SFT-100K
```

You need to agree to share your contact information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Project | Code | Collection

OpenThinkerAgent-32B-SFT-100K

OpenThoughts-Agent is an open-source effort to curate the best datasets for training agents. Our release includes datasets, models and our research codebase.

OpenThinkerAgent-32B-SFT-100K is post-trained from Qwen/Qwen3-32B with full-parameter SFT on the 100,000-example OpenThoughts-Agent-SFT-100K dataset (Top-4 task sources, GLM-4.7-AWQ teacher in the terminus-2 harness, ≥5-turn trace filter).

Homepage: https://www.openthoughts.ai/blog/agent
Repository: https://github.com/open-thoughts/OpenThoughts-Agent

Performance

Evaluated in the terminus-2 harness (pass@1, mean over 3 stochastic re-runs):

Model	Harness	SWE-Bench-Verified-100	OpenThoughts-TBLite	Terminal-Bench 2.0
Qwen/Qwen3-32B	Terminus-2	26.7	13.7	7.5
OpenThinkerAgent-32B-SFT-100K	Terminus-2	55.7	41.3	26.2

Data

The model is trained on OpenThoughts-Agent-SFT-100K: (task, agent-trajectory) pairs from the Top-4 task sources (SWE-Smith, StackExchange-SuperUser, StackExchange-Tezos with synthetic augmentation, IssueTasks). Trajectories are generated by GLM-4.7-AWQ in the terminus-2 harness and filtered to traces with at least 5 model turns.

Training hyperparameters

learning_rate: 4e-05
lr_scheduler_type: cosine, warmup_ratio 0.1
global_batch_size: 96
num_epochs: 5
cutoff_len: 32768
precision: bf16, DeepSpeed ZeRO-3

Citation

@misc{openthoughts-agent,
  author = {Team, OpenThoughts-Agent},
  title = {{OpenThoughts-Agent: Data Recipes for Agentic Models}},
  howpublished = {https://www.openthoughts.ai/blog/agent},
  year = {2026}
}

Downloads last month: 14

Safetensors

Model size

677k params

Tensor type

BF16

Model tree for Artificial-Production-Units/OpenThinkerAgent-32B-SFT-100K

Base model

Qwen/Qwen3-32B

Finetuned

(507)

this model