SmallThinker-3B-Preview
by Tiiny
3.4B params · text-generation · 415 likes · 188 downloads
SmallThinker-3B-Preview is a 3.4B parameter model. At Q4 quantization it requires 2GB of VRAM. It runs comfortably on GeForce RTX 4090 (386 tok/s), GeForce RTX 5090 (579 tok/s), M4 Max 128GB (141 tok/s).
Inference providers
| Provider | $/1M in | $/1M out | Throughput |
|---|---|---|---|
| Featherless |
GPU compatibility
| GPU | VRAM | Q4 Decode | Verdict |
|---|---|---|---|
| GeForce RTX 4090 | 24GB | 386 tok/s | comfortable |
| GeForce RTX 5090 | 32GB | 579 tok/s | comfortable |
| M4 Max 128GB | 128GB | 141 tok/s | comfortable |
| M4 Pro 48GB | 48GB | 70 tok/s | comfortable |
| M4 Pro 24GB | 24GB | 70 tok/s | comfortable |
| A100 PCIe 80 GB | 80GB | 708 tok/s | comfortable |
| H100 SXM5 80 GB | 80GB | 1424 tok/s | comfortable |
| GeForce RTX 3090 | 24GB | 341 tok/s | comfortable |
| Radeon RX 7900 XTX | 24GB | 282 tok/s | comfortable |
| GeForce RTX 4080 | 16GB | 274 tok/s | comfortable |