Strong open embedding model for retrieval. Small and quick.
embeddingsrag · sizes: 335M · Apache-2.0
| machine | best fit | verdict | speed |
|---|---|---|---|
| Apple M1 · 8GB | 335M F16 | ✅ Runs comfortably | 30–50 tok/s |
| Apple M1 Pro · 16GB | 335M F16 | ✅ Runs comfortably | 75–130 tok/s |
| Apple M2 · 16GB | 335M F16 | ✅ Runs comfortably | 40–70 tok/s |
| Apple M3 Pro · 18GB | 335M F16 | ✅ Runs comfortably | 60–100 tok/s |
| Apple M2 Pro · 32GB | 335M F16 | ✅ Runs comfortably | 75–130 tok/s |
| Apple M4 Pro · 48GB | 335M F16 | ✅ Runs comfortably | 100–160 tok/s |
| Apple M3 Max · 36GB | 335M F16 | ✅ Runs comfortably | 130–210 tok/s |
| Apple M2 Max · 64GB | 335M F16 | ✅ Runs comfortably | 130–210 tok/s |
| Apple M4 Max · 128GB | 335M F16 | ✅ Runs comfortably | 160–270 tok/s |
| Apple M2 Ultra · 192GB | 335M F16 | ✅ Runs comfortably | 170–280 tok/s |
| NVIDIA GeForce RTX 3060 · 12GB VRAM | 335M F16 | ✅ Runs comfortably | 110–180 tok/s |
| NVIDIA GeForce RTX 4070 · 12GB VRAM | 335M F16 | ✅ Runs comfortably | 150–250 tok/s |
| NVIDIA GeForce RTX 4090 · 24GB VRAM | 335M F16 | ✅ Runs comfortably | 300–500 tok/s |
| NVIDIA GeForce RTX 5080 · 16GB VRAM | 335M F16 | ✅ Runs comfortably | 280–470 tok/s |
| NVIDIA GeForce RTX 5090 · 32GB VRAM | 335M F16 | ✅ Runs comfortably | 530–880 tok/s |
| AMD Radeon RX 7900 XTX · 24GB VRAM | 335M F16 | ✅ Runs comfortably | 270–450 tok/s |
Shopping for a machine? What hardware do I need to run mxbai-embed-large? →
Check your exact rig: open the checker →