What is Omnara?
Durable agents that survive crashes and resume, behind one API.
Who is Omnara best for?
Omnara is best for developer teams building their own AI agents that need to keep state and wait for human approvals.
Agents inside your own product
A REST API, a TypeScript SDK and a CLI drive the same control plane, so your app can launch, steer and watch agents for its users.
Long agent jobs that must not restart
Agents keep their state and resume after a failed model call or crash instead of starting the task again.
Internal Slack agents
A first-party Slack app lets a team talk to agents and approve risky tool calls in a channel, with every action logged.
Teams that want to pick model and host
It works with OpenAI, Anthropic, OpenRouter, LiteLLM or self-hosted models, and the whole platform can run in your own VPC.
How much does Omnara cost?
Omnara is free to use and to self-host, with no platform or per-seat fee; you pay only if you run Omnara-managed models, at the provider's own token rates, or managed machines, at $0.0414 per GiB of memory per hour.
- Full platform, no per-seat fees
- Your own model keys and machines at no charge
- Managed models at the provider's token rates
Omnara-managed machines cost $0.0414 per GiB of memory per hour while online, plus $0.20016 per GiB of memory per 30 days for keeping them; idle machines sleep and stop runtime charges. The code is open source under Apache 2.0 and free to self-host, with no card needed. Enterprise, priced on request, adds on-prem or your-own-cloud deployment, SSO and audit-log setup, fine-tuning on your data and custom SLAs. Full pricing on omnara.com. We checked these prices on .
How do readers rate Omnara?
No one has rated Omnara yet. Be the first.
Sign in to rate. One rating per account.
Is Omnara open source?
Yes. Omnara's code is public under the Apache 2.0 licence, so you can read it, change it and run it yourself.
On GitHub
3,395stars
+483 (+16.6%) in the last 30 days
- Last commit
- 13 Nov
- Repository age
- 16 months
- Latest release
- 13 Nov
- Licence
- Apache 2.0
- Repository
- omnara-ai/omnara
What are the best Omnara alternatives?
Omnara is an open source alternative to Mistral Workflows and Claude Managed Agents. Agno, eve, NemoClaw and Trigger.dev do the same job and are open source too.
Mistral WorkflowsFreeDurable multi-step AI processes with retries and human approvals.
Claude Managed AgentsPaidHosted agent sessions with memory and orchestration on Anthropic's platform.
AgnoOpen SourcePython framework and runtime for building AI agents and teams and serving them from your own cloud.
eveOpen SourceVercel's open-source framework for durable backend agents.
NemoClawOpen SourceNVIDIA's sandbox for agents that routes credentials and controls network access.
Trigger.devOpen SourceManaged agents and workflows with retries and observability.
Questions about Omnara
What is Omnara?
Durable agents that survive crashes and resume, behind one API.
Who is Omnara for?
Omnara is best for developer teams building their own AI agents that need to keep state and wait for human approvals.
Is Omnara free?
Mostly. Omnara charges no platform or seat fee, and bringing your own model keys and machines costs nothing. You pay only for Omnara-managed models at the provider's token rates and managed machines at $0.0414 per GiB of memory per hour.
Is Omnara open source?
Yes. Omnara's code is public under the Apache 2.0 licence, so you can read it, change it and run it yourself.
What are the best Omnara alternatives?
The closest alternatives on Recaply are Mistral Workflows, Claude Managed Agents, Agno, eve, NemoClaw and Trigger.dev.
Where can I try Omnara?
On its official website, omnara.com.
Can I self-host Omnara?
Yes. Omnara is open source under the Apache 2.0 licence on GitHub, and you can run it in your own VPC at no cost. You can also start on Omnara's hosted cloud and move to self-hosting later; Enterprise adds on-prem deployment help.
AIMLAPI
Featherless
Eden AI
LLMFly AI
treg
LiteLLM