Models › Qwen3-ASR-1.7B
Qwen3-ASR-1.7B: VRAM needs and speed by GPU
No measured runs yet. Speeds below are estimates. Submit a run.
2.03B parameters · 65,536-token context · GGUF published Apr 13, 2026 · GGUF files on Hugging Face
Memory needed per quantization
| Quant | File size | Needs (estimate) | Smallest card it fits |
|---|---|---|---|
| Q8_0 | 2.2 GB | 3.0 GB | 8 GB cards |
| BF16 | 4.1 GB | 4.9 GB | 8 GB cards |
Expected speed on popular hardware
| Hardware | Best quant that fits | Expected tok/s |
|---|---|---|
| NVIDIA RTX 3060 | Q8_0 | 142 estimate |
| NVIDIA RTX 4060 Ti | Q8_0 | 113 estimate |
| NVIDIA RTX 3090 | Q8_0 | 303 estimate |
| NVIDIA RTX 4090 | Q8_0 | 336 estimate |
| NVIDIA RTX 5090 | Q8_0 | 530 estimate |
| AMD Radeon RX 7900 XTX | Q8_0 | 337 estimate |
| Apple M4 Pro (20-core GPU) (64 GB) | Q8_0 | 107 estimate |
| Apple M3 Max (40-core GPU) (64 GB) | Q8_0 | 156 estimate |
| AMD Ryzen AI Max+ 395 (Radeon 8060S) | Q8_0 | 96.0 estimate |
Too big for your card?
Run Qwen3-ASR-1.7B on a rented 24-80 GB GPU by the hour.
Rent a bigger GPU on RunPodAffiliate link: if you sign up, RigBench may earn a referral credit. It never affects rankings or numbers.