Loading…
Loading…
Models / bge-reranker-v2-m3
1 model (~568M) · ↓ 15.7M downloads a month · from $0.08/h on GTX 1660 SUPER right now. Each model's versions do the same job; quantized ones (AWQ, FP8, 4-bit) need less GPU memory and run on cheaper machines.
| Version | Runtime | Weights | GPU memory | Cheapest machine now |
|---|---|---|---|---|
| OriginalBAAI/bge-reranker-v2-m3 | Text Embeddings Inference | 1.1 GB | 3 GB | $0.08/h GTX 1660 SUPER |
| ONNX (onnx-community)onnx-community/bge-reranker-v2-m3-ONNX | ONNX Runtime | 570 MB | 1.9 GB | $0.08/h GTX 1660 SUPER |
| onebrain rerank v1 (onebrain-ai)onebrain-ai/onebrain-rerank-v1 | ONNX Runtime | 570 MB | 1.9 GB | $0.08/h GTX 1660 SUPER |
| onnx (a-ivanovitch)a-ivanovitch/bge-reranker-v2-m3-onnx | ONNX Runtime | 1 MB | 1 GB | $0.08/h GTX 1660 SUPER |
Model details come from Hugging Face and are refreshed daily; prices follow the live offers. GPU memory is GridPup's estimate: the weights, room for an 8,192-token conversation (chat models) and the runtime's own memory. Large models are split across the GPUs of one machine ("4x RTX 4090").