GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
CrowdMind
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
CrowdMind
Model Tree
Qwen/Qwen3.5-9B
Input Modalities
Output Modalities
Supported Functionality
Results for the released SFT checkpoint.
| Domain | Benchmark | Metric | Qwen3.5-9B | Qwen3.5-9b-pro (SFT) |
|---|---|---|---|---|
| Code | SWE Pro | avg@3 | 32.0 | 44.6 |
| General | AutomationBench v1.0.6 | avg@1 | 5.0 | 30.3 |
| General | Terminal Bench 2.1 | avg@1 | 27.0 | 37.1 |
| General | Toolathlon-Verified | avg@1 | 25.9 | 35.2 |
| General | OfficeQA | avg@1 | 9.0 | 19.5 |
| General | JobBench | avg@1 | 2.6 | 18.3 |
These numbers are provided for the released SFT checkpoint and should be interpreted in the context of the benchmark versions and metrics shown above.
Qwen3.5-9b-pro is an SFT checkpoint built from Qwen/Qwen3.5-9B.
The model is intended for workflows where the model needs to do more than produce a simple conversational answer, including:
The model repository includes the tokenizer/processor configuration and the custom chat template used for inference.
This model uses the CrowdMind Qwen3.5 chat template included with the checkpoint.
Do not replace the supplied template with an unrelated generic chat template. The template is responsible for formatting normal conversations, controlling thinking behavior, and inserting the correct multimodal placeholders for image and video inputs.
The template supports:
low reasoning effortmedium reasoning effortxhigh reasoning effortFor multimodal messages, the template maps images to:
text
<|vision_start|><|image_pad|><|vision_end|>and videos to:
text
<|vision_start|><|video_pad|><|vision_end|>Vision inputs should be supplied as structured content items rather than manually placing these special tokens into the user's prompt.
Important: The custom template does not allow image or video content in a system message. Put vision inputs in the user message.
For text generation, use a recent SGLang build with Qwen3.5 support.
Start the server:
bash
sglang serve \ --model-path CrowdMind/Qwen3.5-9b-pro \ --reasoning-parser mimo \ --host 0.0.0.0 \ --port 30000Then query the OpenAI-compatible endpoint:
python
from openai import OpenAI client = OpenAI( base_url="http://localhost:30000/v1", api_key="EMPTY",) response = client.chat.completions.create( model="CrowdMind/Qwen3.5-9b-pro", messages=[ { "role": "user", "content": "What is 15% of 240?" } ], max_tokens=2048, extra_body={ "chat_template_kwargs": { "enable_thinking": True } },) message = response.choices[0].message print("Thinking:", getattr(message, "reasoning_content", "") or "")print("Answer:", message.content or "")The supplied chat template provides explicit controls for reasoning.
Three reasoning-effort levels are supported:
| Setting | Intended behavior |
|---|---|
low | Short, focused reasoning |
medium | Detailed reasoning |
xhigh | More extensive reasoning and verification |
Thinking can also be disabled entirely.
python
response = client.chat.completions.create( model="CrowdMind/Qwen3.5-9b-pro", messages=[ { "role": "user", "content": "Explain why the sky appears blue." } ], max_tokens=2048, extra_body={ "chat_template_kwargs": { "enable_thinking": True, "reasoning_effort": "medium", } },)python
extra_body={ "chat_template_kwargs": { "enable_thinking": True, "reasoning_effort": "low", }}python
extra_body={ "chat_template_kwargs": { "enable_thinking": True, "reasoning_effort": "medium", }}python
extra_body={ "chat_template_kwargs": { "enable_thinking": True, "reasoning_effort": "xhigh", }}python
extra_body={ "chat_template_kwargs": { "enable_thinking": False, }}When thinking is disabled, the chat template creates the appropriate empty thinking block before the assistant response.
The model's chat format supports structured multimodal content.
A user message can contain:
python
[ { "type": "image", "image": "...", }, { "type": "text", "text": "Describe this image.", },]or:
python
[ { "type": "video", "video": "...", }, { "type": "text", "text": "Describe what happens in this video.", },]The media should be supplied to the processor/runtime rather than manually
inserting <|image_pad|> or <|video_pad|> tokens.
With a multimodal runtime that accepts OpenAI-compatible content blocks:
python
from openai import OpenAI client = OpenAI( base_url="http://localhost:30000/v1", api_key="EMPTY",) response = client.chat.completions.create( model="CrowdMind/Qwen3.5-9b-pro", messages=[ { "role": "user", "content": [ { "type": "image_url", "image_url": { "url": "https://example.com/image.jpg" }, }, { "type": "text", "text": "Describe this image in detail.", }, ], } ], max_tokens=1024,) print(response.choices[0].message.content)The exact accepted URL/media transport depends on the serving runtime. When
using Transformers directly, use the model's AutoProcessor and the
repository's structured image content format shown below.
python
import torchfrom transformers import AutoProcessor, AutoModelForImageTextToText MODEL_ID = "CrowdMind/Qwen3.5-9b-pro" processor = AutoProcessor.from_pretrained(MODEL_ID) model = AutoModelForImageTextToText.from_pretrained( MODEL_ID, torch_dtype=torch.bfloat16, device_map="auto",) messages = [ { "role": "user", "content": [ { "type": "image", "image": "/path/to/image.jpg", }, { "type": "text", "text": "What objects are visible in this image?", }, ], }] inputs = processor.apply_chat_template( messages, add_generation_prompt=True, tokenize=True, return_dict=True, return_tensors="pt",) inputs = { key: value.to(model.device) if hasattr(value, "to") else value for key, value in inputs.items()} with torch.inference_mode(): output_ids = model.generate( **inputs, max_new_tokens=512, ) output_ids = output_ids[:, inputs["input_ids"].shape[1]:] answer = processor.batch_decode( output_ids, skip_special_tokens=True,)[0] print(answer)Multiple images can be included in the same user message.
python
messages = [ { "role": "user", "content": [ { "type": "image", "image": "/path/to/image1.jpg", }, { "type": "image", "image": "/path/to/image2.jpg", }, { "type": "text", "text": "Compare these two images and explain the differences.", }, ], }]This is useful for:
The custom chat template also supports video content.
A video message uses:
python
messages = [ { "role": "user", "content": [ { "type": "video", "video": "/path/to/video.mp4", }, { "type": "text", "text": "Describe what happens in this video.", }, ], }]The chat template represents the video internally as:
text
<|vision_start|><|video_pad|><|vision_end|>The processor/runtime is responsible for preparing the actual video input.
python
messages = [ { "role": "user", "content": [ { "type": "video", "video": "/path/to/video.mp4", }, { "type": "text", "text": "Summarize the events in this video chronologically.", }, ], }]python
messages = [ { "role": "user", "content": [ { "type": "video", "video": "/path/to/security_camera.mp4", }, { "type": "text", "text": ( "Analyze the video and describe the important events, " "including what changes over time." ), }, ], }]Video preprocessing support can vary by Transformers/SGLang version and hardware configuration. Use a recent runtime with Qwen3.5 multimodal support.
The content format also allows different modalities to appear together.
python
messages = [ { "role": "user", "content": [ { "type": "image", "image": "/path/to/start.jpg", }, { "type": "video", "video": "/path/to/process.mp4", }, { "type": "text", "text": ( "The image shows the starting state and the video shows " "what happened afterward. Explain the changes." ), }, ], }]This format is useful for workflows such as:
Vision inputs can be combined with the reasoning controls.
For example:
python
inputs = processor.apply_chat_template( messages, add_generation_prompt=True, tokenize=True, return_dict=True, return_tensors="pt", enable_thinking=True, reasoning_effort="medium",)For a shorter response:
python
inputs = processor.apply_chat_template( messages, add_generation_prompt=True, tokenize=True, return_dict=True, return_tensors="pt", enable_thinking=True, reasoning_effort="low",)For more extensive reasoning:
python
inputs = processor.apply_chat_template( messages, add_generation_prompt=True, tokenize=True, return_dict=True, return_tensors="pt", enable_thinking=True, reasoning_effort="xhigh",)The preferred message structure is:
python
{ "type": "text", "text": "Your question here"}python
{ "type": "image", "image": "/path/to/image.jpg"}python
{ "type": "video", "video": "/path/to/video.mp4"}These items can be combined in the content list of a user message.
For example:
python
{ "role": "user", "content": [ { "type": "image", "image": "/path/to/image.jpg" }, { "type": "text", "text": "What is happening here?" } ]}The model is tagged for agentic and tool-use applications.
A typical tool-oriented conversation can be represented as normal structured messages, with the serving framework responsible for the actual tool execution.
Example application flow:
text
User ↓Qwen3.5-9b-pro ↓Tool selection / action ↓External tool ↓Tool result ↓Qwen3.5-9b-pro ↓Final responseFor production agent systems, validate tool calls and arguments before executing external actions.
A reasonable starting point for general generation is:
python
max_tokens=2048For coding and agentic tasks, increase the output budget when the task requires longer tool interactions or code.
For concise responses:
python
enable_thinking=FalseFor lightweight reasoning:
python
enable_thinking=Truereasoning_effort="low"For general reasoning:
python
enable_thinking=Truereasoning_effort="medium"For difficult reasoning tasks:
python
enable_thinking=Truereasoning_effort="xhigh"Actual generation settings should be tuned for the deployment workload.
The checkpoint can be loaded with Transformers using the processor and model configuration supplied by the repository.
python
import torchfrom transformers import ( AutoProcessor, AutoModelForImageTextToText,) MODEL_ID = "CrowdMind/Qwen3.5-9b-pro" processor = AutoProcessor.from_pretrained(MODEL_ID) model = AutoModelForImageTextToText.from_pretrained( MODEL_ID, torch_dtype=torch.bfloat16, device_map="auto",)Use a recent Transformers release with Qwen3.5 support.
If your installed Transformers version does not recognize the model or processor, upgrade Transformers before attempting multimodal inference.
This model is an SFT checkpoint and may produce incorrect, incomplete, or overconfident outputs.
Important limitations include:
This model should be evaluated on the specific workloads and safety requirements of the intended application.
CrowdMind/Qwen3.5-9b-pro is intended for research, experimentation, and
applications involving:
Users are responsible for evaluating the model for their particular deployment and ensuring that its use complies with applicable laws, policies, and organizational requirements.
This model is a fine-tuned version of:
Please refer to the base model repository for the upstream model's architecture, pretraining information, capabilities, and applicable documentation.
If you use this model, please reference the model repository:
text
CrowdMind/Qwen3.5-9b-proThe model was released by CrowdMind and developed by Dustin Loring.
This repository is released under the MIT License.
See the repository license file for the complete license text.
CrowdMind acknowledges the Qwen team and the contributors to the
Qwen/Qwen3.5-9B base model.
This model would not exist without the underlying Qwen3.5 model and the broader open-source ML ecosystem.
sglang serve \ --model-path CrowdMind/Qwen3.5-9b-pro \ --reasoning-parser mimo \ --host 0.0.0.0 \ --port 30000from openai import OpenAI client = OpenAI( base_url="http://localhost:30000/v1", api_key="EMPTY",) response = client.chat.completions.create( model="CrowdMind/Qwen3.5-9b-pro", messages=[ { "role": "user", "content": "What is 15% of 240?" } ], max_tokens=2048, extra_body={ "chat_template_kwargs": { "enable_thinking": True } },) message = response.choices[0].message print("Thinking:", getattr(message, "reasoning_content", "") or "")print("Answer:", message.content or "")response = client.chat.completions.create( model="CrowdMind/Qwen3.5-9b-pro", messages=[ { "role": "user", "content": "Explain why the sky appears blue." } ], max_tokens=2048, extra_body={ "chat_template_kwargs": { "enable_thinking": True, "reasoning_effort": "medium", } },)extra_body={ "chat_template_kwargs": { "enable_thinking": True, "reasoning_effort": "low", }}extra_body={ "chat_template_kwargs": { "enable_thinking": True, "reasoning_effort": "medium", }}extra_body={ "chat_template_kwargs": { "enable_thinking": True, "reasoning_effort": "xhigh", }}extra_body={ "chat_template_kwargs": { "enable_thinking": False, }}[ { "type": "image", "image": "...", }, { "type": "text", "text": "Describe this image.", },][ { "type": "video", "video": "...", }, { "type": "text", "text": "Describe what happens in this video.", },]from openai import OpenAI client = OpenAI( base_url="http://localhost:30000/v1", api_key="EMPTY",) response = client.chat.completions.create( model="CrowdMind/Qwen3.5-9b-pro", messages=[ { "role": "user", "content": [ { "type": "image_url", "image_url": { "url": "https://example.com/image.jpg" }, }, { "type": "text", "text": "Describe this image in detail.", }, ], } ], max_tokens=1024,) print(response.choices[0].message.content)import torchfrom transformers import AutoProcessor, AutoModelForImageTextToText MODEL_ID = "CrowdMind/Qwen3.5-9b-pro" processor = AutoProcessor.from_pretrained(MODEL_ID) model = AutoModelForImageTextToText.from_pretrained( MODEL_ID, torch_dtype=torch.bfloat16, device_map="auto",) messages = [ { "role": "user", "content": [ { "type": "image", "image": "/path/to/image.jpg", }, { "type": "text", "text": "What objects are visible in this image?", }, ], }] inputs = processor.apply_chat_template( messages, add_generation_prompt=True, tokenize=True, return_dict=True, return_tensors="pt",) inputs = { key: value.to(model.device) if hasattr(value, "to") else value for key, value in inputs.items()} with torch.inference_mode(): output_ids = model.generate( **inputs, max_new_tokens=512, ) output_ids = output_ids[:, inputs["input_ids"].shape[1]:] answer = processor.batch_decode( output_ids, skip_special_tokens=True,)[0] print(answer)messages = [ { "role": "user", "content": [ { "type": "image", "image": "/path/to/image1.jpg", }, { "type": "image", "image": "/path/to/image2.jpg", }, { "type": "text", "text": "Compare these two images and explain the differences.", }, ], }]messages = [ { "role": "user", "content": [ { "type": "video", "video": "/path/to/video.mp4", }, { "type": "text", "text": "Describe what happens in this video.", }, ], }]messages = [ { "role": "user", "content": [ { "type": "video", "video": "/path/to/video.mp4", }, { "type": "text", "text": "Summarize the events in this video chronologically.", }, ], }]messages = [ { "role": "user", "content": [ { "type": "video", "video": "/path/to/security_camera.mp4", }, { "type": "text", "text": ( "Analyze the video and describe the important events, " "including what changes over time." ), }, ], }]messages = [ { "role": "user", "content": [ { "type": "image", "image": "/path/to/start.jpg", }, { "type": "video", "video": "/path/to/process.mp4", }, { "type": "text", "text": ( "The image shows the starting state and the video shows " "what happened afterward. Explain the changes." ), }, ], }]inputs = processor.apply_chat_template( messages, add_generation_prompt=True, tokenize=True, return_dict=True, return_tensors="pt", enable_thinking=True, reasoning_effort="medium",)inputs = processor.apply_chat_template( messages, add_generation_prompt=True, tokenize=True, return_dict=True, return_tensors="pt", enable_thinking=True, reasoning_effort="low",)inputs = processor.apply_chat_template( messages, add_generation_prompt=True, tokenize=True, return_dict=True, return_tensors="pt", enable_thinking=True, reasoning_effort="xhigh",){ "type": "text", "text": "Your question here"}{ "type": "image", "image": "/path/to/image.jpg"}{ "type": "video", "video": "/path/to/video.mp4"}{ "role": "user", "content": [ { "type": "image", "image": "/path/to/image.jpg" }, { "type": "text", "text": "What is happening here?" } ]}User ↓Qwen3.5-9b-pro ↓Tool selection / action ↓External tool ↓Tool result ↓Qwen3.5-9b-pro ↓Final responseimport torchfrom transformers import ( AutoProcessor, AutoModelForImageTextToText,) MODEL_ID = "CrowdMind/Qwen3.5-9b-pro" processor = AutoProcessor.from_pretrained(MODEL_ID) model = AutoModelForImageTextToText.from_pretrained( MODEL_ID, torch_dtype=torch.bfloat16, device_map="auto",)