Instructions to use AllSpark-Research/Iris-mini with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use AllSpark-Research/Iris-mini with Transformers:

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

pipe = pipeline("text-generation", model="AllSpark-Research/Iris-mini")
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, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("AllSpark-Research/Iris-mini")
model = AutoModelForMultimodalLM.from_pretrained("AllSpark-Research/Iris-mini", device_map="auto")
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 Settings

vLLM

How to use AllSpark-Research/Iris-mini with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/AllSpark-Research/Iris-mini

SGLang

How to use AllSpark-Research/Iris-mini 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 "AllSpark-Research/Iris-mini" \
    --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": "AllSpark-Research/Iris-mini",
		"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 "AllSpark-Research/Iris-mini" \
        --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": "AllSpark-Research/Iris-mini",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use AllSpark-Research/Iris-mini with Docker Model Runner:
```
docker model run hf.co/AllSpark-Research/Iris-mini
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

Iris-mini

Climbing to the Search Frontier.

🤗 Collection | 💻 GitHub | 🔬 Evaluation Harness

Introduction

Iris-mini is an open-weight search agent post-trained from Qwen3.6-35B-A3B. A capable search agent has to decide what to search, how to read what comes back, when to keep going, and when the evidence it has gathered is enough. Iris is trained for exactly that loop, through alternating rounds of supervised fine-tuning and reinforcement learning.

Its larger sibling is Iris-pro (397B-A17B).


Base	Qwen3.6-35B-A3B
Parameters	35B total / 3B active (256 experts, 8 active)
Layers / hidden	40 / 2048
Context	256K
Precision	bfloat16

Performance

Open-weight systems in the 30–35B range:

Model	Size	BrowseComp	BrowseComp-ZH	DeepSearchQA	HLE
MiroThinker-1.7-mini	30B	67.9	72.3	–	36.4
FORT-Searcher	30B	72.2	75.0	–	–
Apodex-1.0-mini	35B	71.5	80.6	82.2	46.8
Nex-N2-mini	35B	74.1	79.6^r	87.2^r	37.1^r
Agents-A1	35B	75.5	–	–	47.6
XYZ-Aquila-mini	35B	78.8	82.9	89.5	51.1
Iris-mini	35B	82.2	84.8	86.9	52.3

DeepSearchQA is scored with F1, the rest with accuracy; HLE uses the 2,158-question text-only subset. Iris numbers use the discard-all context-management setting; baselines come from their public reports, each under its own context management. ^r reproduced by the XYZ-Aquila team.

Context management

Long-horizon search runs out of context before a hard question is resolved, so a published number belongs to the agent and its harness. We report every benchmark in both regimes, under one tool set, one context limit and one judge.

Setting	BrowseComp	BrowseComp-ZH	DeepSearchQA	HLE
w/o	64.7	72.3	81.0	43.2
retry	–	83.0	89.1	52.0
discard-all	82.2	84.8	86.9	52.3
discard-all + retry	85.9	85.1	89.9	52.4

discard-all resets the conversation to the opening question once the prompt crosses a threshold. retry restarts an episode that ended without a parseable answer, carrying forward a short summary of what was already ruled out. We report discard-all as the headline setting even where adding retry scores higher.

Quickstart

Serve with SGLang:

python -m sglang.launch_server \
  --model-path AllSpark-Research/Iris-mini \
  --served-model-name Iris-mini \
  --port 21234 --tp-size 4 \
  --context-length 262144 \
  --reasoning-parser qwen3 \
  --tool-call-parser qwen3_coder

Then evaluate with Iris-Harness, which reproduces the numbers above:

bash scripts/run_eval.sh \
  --base-url http://127.0.0.1:21234/v1 \
  --llm-config iris-mini \
  --benchmarks "browsecomp:0:1" \
  --context-discard-threshold 131072

The model is trained to call tools through the OpenAI function-calling interface and to wrap its final answer in \boxed{}. Prior reasoning is carried back into the conversation on each turn; the harness's iris-mini config sets this up for you.