Nebius
nebius.com
AI-focused cloud platform with GPU infrastructure.
Quick Information
- Category:
- MLOps & Infrastructure
- Pricing:
- Paid
- Platforms:
- Cloud, Kubernetes, API
- Free Plan:
- No
- API Available:
- Yes
- Last Verified:
- October 2026
What Is Nebius?
Nebius is an AI-focused cloud platform with GPU infrastructure and a managed inference service. The key pricing fact is that Nebius raised on-demand GPU prices from October 1, 2026, so older figures such as $3.85 for the H100 are out of date. Nebius's own compute pricing docs say updated prices apply to virtual machines with B300, B200, H200 and H100 GPUs. Secondary trackers put the post-increase on-demand rates at about $4.50 per GPU-hour for H100, $5.40 for H200 and $8.50 for B200, up roughly 17% to 20% from $3.85, $4.50 and $7.15. These come from price-tracking sites, not the docs page itself, so confirm them before budgeting.
Preemptible instances now use dynamic pricing, with one source quoting floors of $0.79 for H100 and $0.99 for B200, capped just under the on-demand rate and repriced as often as every 15 minutes. Reserving large clusters can reportedly earn discounts of up to 35% below on-demand. Token Factory, the managed inference product renamed from Nebius AI Studio in November 2025, is described by one review as serving more than 60 open models such as Llama, Qwen, DeepSeek, GLM, Kimi and GPT-OSS behind an OpenAI-compatible API, with prices starting at $0.06 per million input tokens. I did not find current per-model Token Factory pricing.
Nebius suits AI teams that want GPU clusters or managed open-model inference from a dedicated AI cloud. Compare CoreWeave, Lambda and Crusoe. Browse the MLOps & Infrastructure tools directory for more.
What Can You Use Nebius For?
GPU training clusters
H100 to B300.
Managed open-model inference
Token Factory.
Nebius Features
- Price increase October 2026 — Roughly 17%-21%.
- Dynamic preemptible pricing — Spot-style.
- Token Factory — 60+ open models reported.
Who Is Nebius For?
- AI labs and startups — Dedicated GPU cloud.
- Developers using open models — OpenAI-compatible API.
Nebius Pros & Limitations
What We Like
- Wide GPU range
- Reserved discounts
- Managed inference option
Limitations
- Prices rose in October 2026
- Figures come from secondary trackers
- Token Factory per-model prices not confirmed
Nebius Pricing
| Plan | Price | Main Features |
|---|---|---|
| On-demand GPUs (from Oct 1, 2026) | H100 about $4.50, H200 $5.40, B200 $8.50 per GPU-hour | Per trackers; prior H100 $3.85, H200 $4.50, B200 $7.15 |
| Preemptible | Dynamic; floors H100 from $0.79, B200 from $0.99 | Capped below on-demand; repriced up to every 15 minutes |
| Reserved clusters | Up to 35% below on-demand | Large commitments |
| Token Factory | From $0.06 per million input tokens | Per-model rates not confirmed |
Pricing was last verified October 2026. Check the official website for the latest pricing.
AI Listing Jungle Ranking
- Global Rank:
- #1164
- Overall Score:
- 24.5/100
- Monthly Users:
- –
- Rank Change:
- ▲ 393
This ranking reflects AI Listing Jungle's own independent scoring methodology and is not an endorsement by Nebius or its company. How Our Rankings Work
Frequently Asked Questions
How much is an H100 on Nebius?
About $4.50 per GPU-hour after October 1, 2026.
Did prices change?
Yes, roughly 17%-21% higher.
What is Token Factory?
Managed inference, formerly Nebius AI Studio.
Is there spot pricing?
Dynamic preemptible pricing.
Our Take
Nebius is a credible AI cloud, now pricier.
Check the docs for live rates.
Nebius Alternatives
Nebius Comparisons
Ready to try Nebius?
Visit the official website to get started.
Visit Official WebsiteAI Listing Jungle is an independent AI tools directory. Product names, logos, and trademarks belong to their respective owners.