jhqxxx
5.6 KiB
5.6 KiB
Supported Models
aha supports a growing collection of state-of-the-art AI models across multiple domains.
Language Model
| Model | Parameters | Description | Use Case | License |
|---|---|---|---|---|
| Qwen3-0.6B | 0.6B | Latest generation | Advanced reasoning | Apache 2.0 |
| MiniCPM4-0.5B | 0.5B | Efficient lightweight | Edge deployment | Apache 2.0 |
Vision & Multimodal
| Model | Parameters | Description | License |
|---|---|---|---|
| Qwen2.5-VL-3B | 3B | Image understanding | Qwen Research License |
| Qwen2.5-VL-7B | 7B | Image understanding | Apache 2.0 |
| Qwen3-VL-2B | 2B | Enhanced multimodal | Apache 2.0 |
| Qwen3-VL-4B | 4B | Enhanced multimodal | Apache 2.0 |
| Qwen3-VL-8B | 8B | Enhanced multimodal | Apache 2.0 |
| Qwen3-VL-32B | 32B | Enhanced multimodal | Apache 2.0 |
| Qwen3.5-0.8B | 0.8B | Native Multimodal | Apache 2.0 |
| Qwen3.5-2B | 2B | Native Multimodal | Apache 2.0 |
| Qwen3.5-4B | 4B | Native Multimodal | Apache 2.0 |
| Qwen3.5-9B | 9B | Native Multimodal | Apache 2.0 |
OCR
| Model | Languages | Type | Strength | License |
|---|---|---|---|---|
| PaddleOCR-VL | Multi | Lightweight | General documents | Apache 2.0 |
| PaddleOCR-VL1.5 | Multi | Lightweight | General documents | Apache 2.0 |
| Hunyuan-OCR | Chinese | Deep learning | Complex layouts | Tencent Hunyuan Community License |
| DeepSeek-OCR | Multi | Scene text | Natural images | MIT |
| GLM-OCR | 8 | Scene text | complex document | MIT |
Speech Recognition (ASR)
| Model | Parameters | Language | Real-time | Speed | License |
|---|---|---|---|---|---|
| Fun-ASR-Nano-2512 | 2G | Chinese/English | Yes | Fast | Not Specified |
| GLM-ASR-Nano-2512 | 4.5G | Chinese/English | Yes | Fast | MIT |
| Qwen3-ASR-0.6B | 0.6B | Chinese/English | Yes | Fast | Apache 2.0 |
| Qwen3-ASR-1.7B | 1.7B | Chinese/English | Yes | Fast | Apache 2.0 |
Audio Generation
| Model | Parameters | Type | Description | License |
|---|---|---|---|---|
| VoxCPM-0.5B | 0.5B | Voice Codec | Neural audio codec | Apache 2.0 |
| VoxCPM1.5 | - | Voice Codec | Enhanced voice generation | Apache 2.0 |
Image Processing
| Model | Type | Description | License |
|---|---|---|---|
| RMBG-2.0 | Background Removal | Remove image backgrounds | CC BY-NC 4.0 |
Model Sources
Models are sourced from:
- Hugging Face - Primary model hub
- ModelScope - Chinese model hub
Adding New Models
See Development Guide for instructions on adding new model integrations.
License Compliance
Important: Each model has its own license. Please review the model's license before use in production. Some key considerations:
- Apache 2.0: Permissive, commercial-friendly
- MIT: Permissive, commercial-friendly
- Qwen Research License: Research use, may have restrictions
- Tencent Hunyuan Community License: Custom license, review terms
- CC BY-NC 4.0: Non-commercial only
Always verify license terms before deployment in production environments.
Model Updates
Models updated from time to time.
Performance Benchmarks
Approximate inference speeds on CPU (M1 Pro):
| Model | Task | Tokens/sec |
|---|---|---|
| Qwen3-0.6B | Text | 40-50 |
| Qwen2.5-VL-3B | Vision | 20-30 |
| Qwen3-ASR-0.6B | ASR | 200-500x |
| VoxCPM-0.5B | TTS | Real-time |
Benchmarks vary by hardware and input size.