OLMoE models
4 OLMoE models from Ai2 on Hugging Face, from 6.9B to 6.9B parameters, published by Ai2 in F32 and BF16, with a 4K-token context. The smallest official model, OLMoE-1B-7B-0125-Instruct, needs about 15.5 GB of VRAM at its published precision; the cheapest live fit is RTX A4000 at $0.173/hr.
Part of the OLMo series
Pick a size
One row per official OLMoE size: the VRAM it needs at each precision, the cheapest GPU that holds it today, and the KV cache for a 32K-token context.
| Model | Parameters | Native VRAM | FP8 VRAM | INT4 VRAM | Live GPU fit (native) | Est. $/hr | KV cache at 32K |
|---|---|---|---|---|---|---|---|
| OLMoE-1B-7B-0125-Instruct | 6.9B | 15.5 GB | 7.7 GB | 3.9 GB | RTX A4000 | $0.173/hr | 4.0 GB |
VRAM is weights times a flat 1.2 overhead; the KV cache is a separate per-model figure at 16-bit, shown where the architecture is published. See the methodology.
Official models (4)
| Model | Parameters | Published as | VRAM needed | Live GPU fit | Est. $/hr |
|---|---|---|---|---|---|
| OLMoE-1B-7B-0125 | 6.9B | F32 | 30.9 GB | V100 | $0.220/hr |
| OLMoE-1B-7B-0125-Instruct | 6.9B | BF16 | 15.5 GB | RTX A4000 | $0.173/hr |
| OLMoE-1B-7B-0924 | 6.9B | BF16 | 15.5 GB | RTX A4000 | $0.173/hr |
| OLMoE-1B-7B-0924-Instruct | 6.9B | BF16 | 15.5 GB | RTX A4000 | $0.173/hr |
VRAM is for the precision the model is published in: the weights times a flat 1.2 overhead, with the KV cache not included. See the methodology. The fit shown is the lowest-priced single GPU type that holds the model at that precision, or the lowest-priced multi-GPU set (up to 8) when none does.
More from Ai2
Related on Aquanode: GPU requirements for OLMoE-1B-7B-0125-Instruct.