SV Premium

LLM APIs

Nebius Token Factory

Nebius inference API (formerly AI Studio) for open-weight models, with endpoints in EU, UK, US and other regions.

European teams that want open-weight models with a DPA from an EU-based company can sign with Nebius B.V. and use endpoints in Finland and other EU regions. Residency depends on the model, though: gpt-oss-120b is served in eu-north1 and Kimi-K3 in eu-west2, while DeepSeek-V4-Flash and GLM-5.3 are listed in US and UK regions only. Prices on shared models are not the lowest, with DeepSeek-V4-Pro at $1.75 / $3.50 against $1.30 / $2.60 on DeepInfra.

Headquarters
🇳🇱 Netherlands
Parent jurisdiction
🇳🇱 Netherlands (parent: Nebius Group)
EU self-serve plan
Available; check eligible plans and regions
GDPR DPA
Yes
Billing
payg
Open source
No
Website
nebius.com
Source last checked
2026-10-06
Pricing evidence
Published rates

Pricing

Calculated from the listed vendor rates; exclusions still apply. How to read these estimates.

ModelInput / 1M tokensOutput / 1M tokensCached / 1M tokensContextDetails
Kimi-K3$3$15—1024kfrontieropen weights
GLM-5.3$1.4$4.4—1024kfrontieropen weights
DeepSeek-V4-Pro$1.75$3.5—1000kfrontieropen weights
DeepSeek-V4-Flash-0731$0.14$0.28—1024kmidopen weights
gpt-oss-120b$0.15$0.6—131kmidopen weights

Prices from the official public model catalog (tokenfactory.nebius.com/api/public/models_info), base flavor. Regional endpoints include eu-north1 (Finland) and eu-west1/eu-west2, but each model is served in specific regions: e.g. gpt-oss-120b is in eu-north1, Kimi-K3 in eu-west2, while DeepSeek-V4-Flash and GLM-5.3 are listed in US/UK regions only. DPA with Nebius B.V. at docs.tokenfactory.nebius.com/legal/dpa.

Managed alternatives

At 500M input tokens / month, 50M output tokens / month. Adjust usage and requirements.

Nebius Token Factory · DeepSeek-V4-Flash-0731: $1,008/yr estimated.

The price difference leaves out migration costs, operating costs and any features you’d need to replace. We compare the selected model or the cheapest listed model in the chosen capability tier. Sharing a tier does not mean equal quality.

Found a wrong rate or plan restriction? Report a correction with a current source.