Whisper models
30 Whisper models from OpenAI on Hugging Face, from 242M to 1.6B parameters, published by OpenAI in F32 and F16. The smallest official model, whisper-small.en, needs about 1.1 GB of VRAM at its published precision; the cheapest live fit is RTX 3070 at $0.088/hr.
Pick a size
One row per official Whisper 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 |
|---|---|---|---|---|---|---|---|
| whisper-small | 242M | 1.1 GB | 0.3 GB | 0.1 GB | RTX 3070 | $0.088/hr | not published for this architecture |
| whisper-large-v3-turbo | 809M | 1.8 GB | 0.9 GB | 0.5 GB | RTX 3070 | $0.088/hr | not published for this architecture |
| whisper-large-v3 | 1.5B | 3.4 GB | 1.7 GB | 0.9 GB | RTX 3070 | $0.088/hr | not published for this architecture |
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 (8)
| Model | Parameters | Published as | VRAM needed | Live GPU fit | Est. $/hr |
|---|---|---|---|---|---|
| whisper-small.en | 242M | F32 | 1.1 GB | RTX 3070 | $0.088/hr |
| whisper-small | 242M | F32 | 1.1 GB | RTX 3070 | $0.088/hr |
| whisper-medium.en | 764M | F32 | 3.4 GB | RTX 3070 | $0.088/hr |
| whisper-medium | 764M | F32 | 3.4 GB | RTX 3070 | $0.088/hr |
| whisper-large-v3-turbo | 809M | F16 | 1.8 GB | RTX 3070 | $0.088/hr |
| whisper-large | 1.5B | F32 | 6.9 GB | RTX 3070 | $0.088/hr |
| whisper-large-v2 | 1.5B | F32 | 6.9 GB | RTX 3070 | $0.088/hr |
| whisper-large-v3 | 1.5B | F16 | 3.4 GB | RTX 3070 | $0.088/hr |
Fine-tunes and community models (22)
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 OpenAI
- gpt-oss (6 models)