Nebius
Nebius is an AI cloud company listed on Nasdaq. It rents NVIDIA GPU servers by the hour, from single GPUs to clusters joined by InfiniBand, in Finland, France, Spain, the UK, the US and Israel, and sells open-weight models per token through its Token Factory API.
- Website: https://nebius.com
- Headquarters: Amsterdam, Netherlands
- Kind: GPU cloud, Per-token API
Nebius sells 17 of the compared models per token and prices 6 kinds of GPU for hosting your own.
Models sold per token
| Model | Input, per 1M tokens | Output, per 1M tokens |
|---|---|---|
| DeepSeek V4 Flash 0731 | $0.14 | $0.28 |
| DeepSeek V4 Pro | $1.75 | $3.50 |
| DeepSeek V4 Pro 0813 | $1.32 | $3.96 |
| Gemma 3 27B | $0.10 | $0.30 |
| GLM 5.1 | $1.40 | $4.40 |
| GLM 5.2 | $1.40 | $4.40 |
| Hermes 4 405B | $1.00 | $3.00 |
| Kimi K2.6 | $0.95 | $4.00 |
| Kimi K2.7 Code | $0.95 | $4.00 |
| Kimi K3 | $3.00 | $15.00 |
| MiniCPM-V 4.5 | $0.658 | $1.11 |
| MiniMax M3 | $0.30 | $1.20 |
| OpenAI gpt-oss 120B | $0.15 | $0.60 |
| Qwen3 235B A22B Instruct 2507 | $0.20 | $0.60 |
| Qwen3 30B A3B Instruct 2507 | $0.10 | $0.30 |
| Qwen3.5 397B A17B | $0.60 | $3.60 |
| Qwen3.8 27B | $0.45 | $3.00 |
GPUs for hosting your own
| GPU | From, per GPU-hour | Cheapest instance | GPUs in it |
|---|---|---|---|
| L40S | $1.55 | gpu-l40s-a.1gpu-8vcpu-32gb | 1 × 48 GB |
| RTX PRO 6000 | $1.80 | gpu-rtx6000-a.1gpu-24vcpu-218gb | 1 × 96 GB |
| H100 | $4.50 | gpu-h100-sxm.1gpu-16vcpu-200gb | 1 × 80 GB |
| H200 | $5.40 | gpu-h200-sxm.1gpu-16vcpu-200gb | 1 × 141 GB |
| B200 | $8.50 | gpu-b200-sxm-a.1gpu-20vcpu-224gb | 1 × 180 GB |
| B300 | $9.50 | gpu-b300-sxm.1gpu-24vcpu-346gb | 1 × 288 GB |
Where it runs
9 regions in 6 countries.
| Country | Regions | Per-token models | GPU kinds |
|---|---|---|---|
| Finland | eu-north1 | 7 | 3 |
| United States | us-central1, us-north1 | 6 | 4 |
| United Kingdom | uk-south1, uk-south2 | 3 | 2 |
| France | eu-west1, eu-west2 | 1 | 2 |
| Israel | me-west1 | — | 1 |
| Spain | eu-south1 | — | 1 |
Where the prices come from
- Per-token prices: tokenfactory.nebius.com/api/public/models_info
- GPU servers: docs.nebius.com/compute/resources/pricing.md
- Data sent out (egress): nebius.com/prices
Sources
List prices collected 30 October 2026, refreshed every three hours. No negotiated rates or taxes.