Artificial intelligence news.
Straight from the original announcements.

Kilo adds three ways to use Nebius Token Factory inference

September 30, 2026, 21:01Source: Kilo Code
Kilo adds three ways to use Nebius Token Factory inference
Image: Kilo Code

Kilo, the AI coding tool, now lets developers use Nebius Token Factory inference in three ways: with a personal key on their own machine, with a shared key in the cloud, or through an organization-wide allowlist in Kilo Gateway. The company laid out the options in a tutorial published on its blog.

Key points
  • Kilo describes three ways to use Nebius Token Factory: local BYOK, cloud BYOK and a Kilo Gateway allowlist.
  • Local setup in VS Code goes through Settings > Providers, and the CLI uses the /connect command.
  • A key added in Kilo Cloud can be used by Cloud Agents and shared with a team through a usage pool.
  • Kilo Gateway supports 500+ models from 60+ providers, and admins can restrict Nebius models and regions.
  • The Nebius x NVIDIA global AI hackathon is still open, according to Kilo.

Background: Builders & Brews and the Nebius hackathon

According to Kilo, the company appeared at two Builders & Brews events last Friday. Job Rietbergen spoke in Amsterdam and Brian Turcotte spoke in New York City. The lineup on stage also included JetBrains, LangChain, Nebius and Tavily.

Builders & Brews is a running series from Nebius and Tavily held in cities worldwide. Kilo says these events were part of the Nebius x NVIDIA global AI hackathon, which offers prizes. The hackathon is still open, and Kilo invites interested developers to sign up, get a Nebius Token Factory key and start building in Kilo.

Option one and two: bring your own key

The quickest route is to take a key from Nebius Token Factory and add it on a local machine. Users first need to install Kilo, either in the CLI or in their preferred IDE. In Kilo for VS Code, the path is Settings > Providers > Show more providers > Nebius Token Factory > Add API key. Kilo says the flow for JetBrains and the CLI is similar, and in the CLI the key can be connected with the /connect command.

Once the key is added, users open the model picker and select any model tagged Nebius Token Factory. At that point, in Kilo's words, "you're running on Nebius."

The second option uses the same key in the cloud. Users add it under Bring Your Own Key (BYOK) in Kilo Cloud. Cloud Agents can use the key as well. Team members in Kilo can add the key to the organization so that others can draw from a shared usage pool.

Option three: an org-wide allowlist in Kilo Gateway

The third path runs through Kilo Gateway, which Kilo says supports 500+ models from 60+ providers out of the box. To include Nebius inference, an administrator goes to Settings > Providers, enables Nebius Token Factory, and selects which models and regions the team may use.

According to Kilo, this lets a team route requests through approved models and provider routes. The accompanying screenshot shows the provider allowlist with Nebius Token Factory enabled, alongside region and training-data filters.

What it means for developers and teams

The three setups cover different needs. An individual developer can start with a personal key in minutes. A team can pool usage under one shared key in Kilo Cloud. An organization that wants central control can limit which Nebius models and regions are available through the gateway allowlist.

Kilo points developers to its docs for details and to the Builders & Brews schedule for upcoming events near them.

KiloNebiusNebius Token FactoryKilo CloudKilo Gatewaycodingcloudpartnershipsenterprise AI

Related news

Langfuse launches open source AI Gateway in public beta
Products and toolsNovember 9, 2026

Langfuse launches open source AI Gateway in public beta

Langfuse has launched its AI Gateway in public beta, offering companies a central entry point for model traffic, usage tracking and access governance. The MIT-licensed…

Mistral expands Studio with managed agents, workflows and apps
Products and toolsNovember 5, 2026, 13:00

Mistral expands Studio with managed agents, workflows and apps

Mistral AI is expanding its Mistral Studio enterprise AI platform with managed agents and apps, plus new capabilities for workflows, enterprise context, observability…

Business
BusinessNovember 5, 2026, 08:01

Sinch Q3 sales rise 5% organically as Americas drives growth and AI ties deepen

Swedish cloud communications company Sinch reported third quarter 2026 net sales of SEK 7,085m, an organic increase of 5 percent, with the Americas region cited as the…