nebius
nebius
19 models on nebius, with the price per million tokens and who ends up holding your prompt.
- Company based in
- 🇳🇱 Netherlands
- How it serves
- Runs 19 models itself
- Training on your prompts
- no training
- Data policy source
- docs.tokenfactory.nebius.com
- Policy checked
- 2026-09-26
Models
| Model | Input $/M | Output $/M | Blended* | Holds your prompt | Training |
|---|---|---|---|---|---|
| DeepSeek V4 Flash 0731 | $0.14 | $0.28 | $0.18 | runs it 🇳🇱 Netherlands | no training |
| DeepSeek V4 Pro | $1.75 | $3.50 | $2.19 | runs it 🇳🇱 Netherlands | no training |
| DeepSeek V4 Pro 0813 | $1.32 | $3.96 | $1.98 | runs it 🇳🇱 Netherlands | no training |
| DeepSeek V4.1 Flash | $0.30 | $1.20 | $0.52 | runs it 🇳🇱 Netherlands | no training |
| GLM-5.1 | $1.40 | $4.40 | $2.15 | runs it 🇳🇱 Netherlands | no training |
| GLM-5.2 | $1.40 | $4.40 | $2.15 | runs it 🇳🇱 Netherlands | no training |
| GLM-5.3 | $1.40 | $4.40 | $2.15 | runs it 🇳🇱 Netherlands | no training |
| GLM-5.3-Flash | $0.15 | $0.50 | $0.24 | runs it 🇳🇱 Netherlands | no training |
| GPT OSS 120B | $0.15 | $0.60 | $0.26 | runs it 🇳🇱 Netherlands | no training |
| Kimi K2.6 | $0.95 | $4.00 | $1.71 | runs it 🇳🇱 Netherlands | no training |
| Kimi K2.7 Code | $0.95 | $4.00 | $1.71 | runs it 🇳🇱 Netherlands | no training |
| Kimi K3 | $3.00 | $15.00 | $6.00 | runs it 🇳🇱 Netherlands | no training |
| MiniMax-M3 | $0.30 | $1.20 | $0.52 | runs it 🇳🇱 Netherlands | no training |
| Muse Spark Open | $1.25 | $4.25 | $2.00 | runs it 🇳🇱 Netherlands | no training |
| Nemotron 3 Super 120B A12B | $0.30 | $0.90 | $0.45 | runs it 🇳🇱 Netherlands | no training |
| Nemotron 3 Ultra 550B A55B | $1.00 | $3.00 | $1.50 | runs it 🇳🇱 Netherlands | no training |
| Nemotron 3.5 Lightning 30B A3B | $0.06 | $0.24 | $0.10 | runs it 🇳🇱 Netherlands | no training |
| NVIDIA Nemotron 3.5 Super VL | $0.30 | $0.90 | $0.45 | runs it 🇳🇱 Netherlands | no training |
| Qwen3.8 27B | $0.45 | $3.00 | $1.09 | runs it 🇳🇱 Netherlands | no training |
Blended is our calculation: (3 × input + output) ÷ 4, US dollars per million tokens — most calls read far more than they write. Tiers that are cheaper because they train on your data are not used for the “from” price.