svara-tts-v1: GPU memory and hosting cost
Open weights · 3.3 billion parameters · 131,072-token context · licence apache-2.0
No cloud sells this model per token: to use it you run it yourself, on GPU instances you rent.
Run it yourself
It needs 16.5 GB of GPU memory at fp32, the precision it ships in.
| Cloud | Cheapest instance that fits | GPUs | An hour | A month, around the clock |
|---|---|---|---|---|
| Google Cloud | g2-standard-4 | 1 × L4 24 GB | $0.7068 | $516 |
| AWS | g6.xlarge | 1 × L4 24 GB | $0.8048 | $588 |
| Azure | Standard_NC36ds_xl_RTXPRO6000BSE_v6 | 1 × RTX PRO 6000 24 GB | $1.44 | $1,053 |
| Alibaba Cloud | ecs.gn7i-c8g1.2xlarge | 1 × A10 24 GB | $1.68 | $1,224 |
More from Other
- dolphin-2.9.1-yi-1.5-34b
- JiRackUltra_1b
- OTel-2.0-LLM-31B-IT
- moondream2
- TinyLlama-1.1B-Chat-v1.0
- MiniCPM5-2B
Similar size, open weights
- Qwen3-4B
- Qwen2.5-3B-Instruct (Qwen)
- Qwen3-4B-Instruct-2507 (Qwen)
- Qwen3-1.7B (Qwen)
- NVIDIA-Nemotron-3-Nano-4B-BF16
- Qwen3-Reranker-4B
List prices collected 5 October 2026, refreshed every three hours. No negotiated rates or taxes.