Loading…
Loading…
Models / MiroThinker-v1.5
GPU memory needed (smallest version of each): MiroThinker-v1.5 30B about 71 GB, MiroThinker-v1.5 235B about 526.4 GB. The cheapest runs on 2x Nvidia A100 from $1.09/h right now.
2 models (30B to 235B) · ↓ 520 downloads a month · from $1.09/h on 2x Nvidia A100 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 |
|---|---|---|---|---|
| Originalmiromind-ai/MiroThinker-v1.5-30B | vLLM | 61 GB | 71 GB | $1.09/h 2x Nvidia A100 |
| Version | Runtime | Weights | GPU memory | Cheapest machine now |
|---|---|---|---|---|
| Originalmiromind-ai/MiroThinker-v1.5-235B | vLLM | 470 GB | 526.4 GB | $13.98/h 8x RTX PRO 6000 |
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").