phi-1_5: GPU memory and hosting cost
Open weights · 1.42 billion parameters · 2,048-token context · licence mit
No cloud sells this model per token: to use it you run it yourself, on GPU instances you rent.
Run it yourself
It needs 3.5 GB of GPU memory at fp16, the precision it ships in.
| Cloud | Cheapest instance that fits | GPUs | An hour | A month, around the clock |
|---|---|---|---|---|
| Azure | Standard_NV6ads_A10_v5 | 1 × A10 4 GB | $0.454 | $331 |
| AWS | g4dn.xlarge | 1 × T4 16 GB | $0.526 | $384 |
| Google Cloud | g2-standard-4 | 1 × L4 24 GB | $0.7068 | $516 |
| Alibaba Cloud | ecs.gn6i-c4g1.xlarge | 1 × T4 16 GB | $1.12 | $816 |
More from Microsoft
Similar size, open weights
- Qwen3-0.6B (Qwen)
- Llama-3.2-1B-Instruct (meta-llama)
- Qwen2.5-1.5B-Instruct (Qwen)
- whisper-large-v3-turbo
- gemma-3-1b-it (google)
- Qwen3-1.7B (Qwen)
List prices collected 5 October 2026, refreshed every three hours. No negotiated rates or taxes.