You can deploy this model and start earning money today!

qualcomm/
Qwen3.5-0.8B

Name: Qwen3.5-0.8B
Availability: InStock

Views

No views yet

Qwen3.5-0.8B: Optimized for Qualcomm Devices

Qwen3.5 is the latest multilingual language model series from Alibaba Cloud with improved reasoning and instruction-following capabilities over Qwen3.

This is based on the implementation of Qwen3.5-0.8B found here. This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the Qualcomm® AI Hub Models library to export with custom configurations. More details on model performance across various devices, can be found here.

Qualcomm AI Hub Models uses Qualcomm AI Hub Workbench to compile, profile, and evaluate this model. Sign up to run these models on a hosted Qualcomm® device.

Deploying Qwen3.5-0.8B on-device

Follow the GenieX quickstart to install GenieX and deploy the model on a target device.

Getting Started

There are two ways to deploy this model on your device:

Option 1: Download Pre-Exported Models

Below are pre-exported model assets ready for deployment.

Runtime	Precision	Chipset	SDK Versions	Download
GENIEX_LLAMACPP	q4_0	Universal		Download

For more device-specific assets and performance metrics, visit Qwen3.5-0.8B on Qualcomm® AI Hub.

Option 2: Export with Custom Configurations

Use the Qualcomm® AI Hub Models Python library to compile and export the model with your own:

Custom weights (e.g., fine-tuned checkpoints)
Custom input shapes
Target device and runtime configurations

This option is ideal if you need to customize the model beyond the default configuration provided here.

See our repository for Qwen3.5-0.8B on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.text_generation

Model Stats:

Model architecture: Transformer with GQA and SwiGLU
Response Rate: Rate of response generation after the first response token.
Supported languages: 100+ languages and dialects
TTFT: Time To First Token is the time it takes to generate the first response token. This is expressed as a range because it varies based on the length of the prompt.

Performance Summary

Model	Runtime	Precision	Chipset	Context Length	Response Rate (tokens per second)	Time To First Token (range, seconds)
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Snapdragon® 8 Elite Gen 5 Mobile	512	75.035642	0.820253 - 3.281012
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Snapdragon® 8 Elite Gen 5 Mobile	512	36.508582	0.489328 - 1.957312
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Snapdragon® 8 Elite Gen 5 Mobile	512	36.840554	0.402922 - 1.611688
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Snapdragon® 8 Elite Gen 5 Mobile	4096	55.71341	11.875686 - 380.021952
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Snapdragon® 8 Elite Gen 5 Mobile	4096	29.304888	4.37993 - 140.15776
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Snapdragon® 8 Elite Gen 5 Mobile	4096	34.335943	3.49085 - 111.7072
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Snapdragon® 8 Elite Mobile	512	72.950102	0.26924750000000003 - 1.0769900000000001
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Snapdragon® 8 Elite Mobile	512	62.441461	0.26454675 - 1.058187
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Snapdragon® 8 Elite Mobile	512	37.52486	0.1532055 - 0.612822
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Snapdragon® 8 Elite Mobile	4096	49.751244	0.42967528125 - 13.749609
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Snapdragon® 8 Elite Mobile	4096	48.911714	0.43504246875 - 13.921359
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Snapdragon® 8 Elite Mobile	4096	35.893754	0.1591485625 - 5.092754
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Snapdragon® X2 Elite	512	90.934149	0.36497199999999996 - 1.4598879999999999
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Snapdragon® X2 Elite	512	62.435826	0.246797 - 0.987188
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Snapdragon® X2 Elite	512	32.242853	0.39581099999999997 - 1.5832439999999999
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Snapdragon® X2 Elite	4096	82.572548	4.599211 - 147.174752
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Snapdragon® X2 Elite	4096	63.42337	2.265519 - 72.496608
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Snapdragon® X2 Elite	4096	30.128671	3.29479 - 105.43328
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Snapdragon® X Elite	512	64.713531	0.12653075 - 0.506123
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Snapdragon® X Elite	512	63.044187	0.12997225 - 0.519889
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Snapdragon® X Elite	512	29.988464	0.2527895 - 1.011158
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Snapdragon® X Elite	4096	51.537779	0.21393696875 - 6.845983
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Snapdragon® X Elite	4096	48.821458	0.23768365625 - 7.605877
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Snapdragon® X Elite	4096	28.38794	0.25720621875 - 8.230599
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Qualcomm® Dragonwing™ IQ-8275	512	24.26831	3.851939 - 15.407756
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Qualcomm® Dragonwing™ IQ-8275	512	12.065672	5.186901 - 20.747604
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Qualcomm® Dragonwing™ IQ-8275	512	12.122275	1.467962 - 5.871848
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Qualcomm® Dragonwing™ IQ-8275	4096	12.796397	37.273388 - 1192.748416
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Qualcomm® Dragonwing™ IQ-8275	4096	9.174312	49.459292999999995 - 1582.6973759999998
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Qualcomm® Dragonwing™ IQ-8275	4096	11.618693	11.452279 - 366.472928
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Qualcomm® Dragonwing™ IQ-9075	512	50.507601	1.442949 - 5.771796
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Qualcomm® Dragonwing™ IQ-9075	512	21.022536	1.82265 - 7.2906
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Qualcomm® Dragonwing™ IQ-9075	512	9.745996	1.309253 - 5.237012
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Qualcomm® Dragonwing™ IQ-X7181	512	64.713531	0.12653075 - 0.506123
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Qualcomm® Dragonwing™ IQ-X7181	512	63.044187	0.12997225 - 0.519889
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Qualcomm® Dragonwing™ IQ-X7181	512	29.988464	0.2527895 - 1.011158
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Qualcomm® Dragonwing™ Q-8750	512	72.950102	0.26924750000000003 - 1.0769900000000001
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Qualcomm® Dragonwing™ Q-8750	512	62.441461	0.26454675 - 1.058187
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Qualcomm® Dragonwing™ Q-8750	512	37.52486	0.1532055 - 0.612822
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Qualcomm® Dragonwing™ IQ-9075	4096	38.005473	14.725108 - 471.203456
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Qualcomm® Dragonwing™ IQ-9075	4096	18.64628	17.464153999999997 - 558.8529279999999
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Qualcomm® Dragonwing™ IQ-9075	4096	9.76354	10.305826 - 329.786432
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Qualcomm® Dragonwing™ IQ-X7181	4096	51.537779	0.21393696875 - 6.845983
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Qualcomm® Dragonwing™ IQ-X7181	4096	48.821458	0.23768365625 - 7.605877
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Qualcomm® Dragonwing™ IQ-X7181	4096	28.38794	0.25720621875 - 8.230599
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Qualcomm® Dragonwing™ Q-8750	4096	49.751244	0.42967528125 - 13.749609
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Qualcomm® Dragonwing™ Q-8750	4096	48.911714	0.43504246875 - 13.921359
Qwen3.5-0.8B	GENIEX_LLAMACPP	q4_0	Qualcomm® Dragonwing™ Q-8750	4096	35.893754	0.1591485625 - 5.092754

License

The license for the original implementation of Qwen3.5-0.8B can be found here.

References

Community

Join our AI Hub Slack community to collaborate, post questions and learn more about on-device AI.
For questions or feedback please reach out to us.

Usage and Limitations

This model may not be used for or in connection with any of the following applications:

Accessing essential private and public services and benefits;
Administration of justice and democratic processes;
Assessing or recognizing the emotional state of a person;
Biometric and biometrics-based systems, including categorization of persons based on sensitive characteristics;
Education and vocational training;
Employment and workers management;
Exploitation of the vulnerabilities of persons resulting in harmful behavior;
General purpose social scoring;
Law enforcement;
Management and operation of critical infrastructure;
Migration, asylum and border control management;
Predictive policing;
Real-time remote biometric identification in public spaces;
Recommender systems of social media platforms;
Scraping of facial images (from the internet or otherwise); and/or
Subliminal manipulation

qualcomm/Qwen3.5-0.8B