All comparisons

Felan vs. Ona

Both run coding agents in the background, from schedules, pull requests, and webhooks. Ona is now part of OpenAI and runs Codex on OpenAI models. Felan runs models from every provider, including local and private ones, with no markup on tokens.

Get startedContact sales

At a Glance

  • Agent

    Felan: Open-source Felan Code, tuned for token cost

    Ona: Codex, OpenAI's open-source agent

  • Models

    Felan: All major providers, plus local and private models

    Ona: OpenAI models; other providers by exception on Enterprise

  • Your own keys

    Felan: Bring your own keys or subscription

    Ona: ChatGPT plan on Core; provider keys by exception on Enterprise

  • Managed tokens

    Felan: Felan credits at provider rates, no markup

    Ona: Billed in OCUs; per-token rates not published

  • Triggers

    Felan: Events, schedules, or by hand

    Ona: Pull requests, schedules, webhooks, or by hand

  • Self-hosting

    Felan: Whole platform on your infrastructure, on Enterprise

    Ona: Ona-managed VPC in your AWS or GCP account, on Enterprise

  • Team plan

    Felan: €99 per month, unlimited members

    Ona: From $20 per month, up to 100 members

Where Felan Differs

Models From Every Provider

Ona Cloud offers OpenAI models only since Anthropic access ended. Felan runs every major provider, plus local and private models.

  • Anthropic
  • OpenAI
  • Google
  • DeepSeek
  • Moonshot AI
  • Local model
  • Private endpoint

No Markup on Tokens

Ona bills agent work in OCUs, at an approximate cost per task. Felan credits are billed at your provider's per-token rates.

Model tokens
Provider rate
Felan markup
None
You pay
Your provider’s price

Self-Host on Your Terms

Ona Enterprise runs in your AWS or GCP account, managed by Ona. Felan Enterprise runs the whole platform on your infrastructure, with local models if you need them.

Your infrastructure

  • Agent loop
  • Tools
  • Repos
  • Model endpoint

Common Questions

Is Felan an Ona alternative?

Yes. Both run coding agents in the background and open pull requests for review. Felan runs models from every provider, including local and private models, and bills tokens at provider rates.

Is Ona part of OpenAI?

Yes. Gitpod became Ona in 2025, and OpenAI's acquisition of Ona closed on August 10, 2026. Ona is now part of the Codex team.

Which models does Ona support?

Ona Cloud runs Codex on OpenAI models. Ona-managed Anthropic access ended, and other token providers are available by exception on Enterprise. You can still install Claude Code in an Ona environment and pay Anthropic directly.

How does Ona billing work?

Core plans start at $20 a month and include OCUs; add-ons cost $10 per 40 OCUs. A standard environment uses 1 OCU an hour, and agent tasks use roughly 1 to 8 OCUs each. With a connected ChatGPT plan, Codex model use is not charged in OCUs.

When is Ona a better fit?

When your team has standardized on Codex and OpenAI models, or wants governed cloud development environments for people as well as agents.

Can I self-host Felan and Ona?

Both, on Enterprise. Ona deploys a single-tenant, Ona-managed setup in your AWS or GCP account. Felan Enterprise runs the whole platform on your infrastructure.

Sources: Ona pricing, Ona Codex Agent, Ona model access, Ona automations, Ona Cloud, Ona cost and budgets, Ona is joining OpenAI, Anthropic model access in Ona, Gitpod is now Ona.

Keep Your Attention for the Hard Problems

Start with one repo and one recurring task.

Get startedContact sales