Implemented for you. Owned by you.
Speculos engineers set up the open-source platform inside your infrastructure and stay until your teams are shipping. The software is free - the fee covers the implementation.
What an implementation includes.
A team of AI-native engineers, working inside your walls until your builders do not need us.
Scope
Sit down with your teams and decide where the platform runs - your cloud, or on-prem.
Set up
Deploy the platform inside your infrastructure - behind your SSO, on your keys, with the models you choose - and wire in your data sources, instructions, and design system.
Hand over
Train your builders, ship the first apps together, and hand you the keys. You run it. We stay on call.
Modules in closed beta.
The open-source core is complete on its own. These commercial modules plug into the same deployment.
Prompt log
every prompt recorded and replayable
Connector catalog
governed connectors beyond the Postgres and MCP references - warehouses, CRMs, and internal systems, granted per person by an admin
Dynamic model routing
the agent picks the model per task; an explicit user pick wins
Common questions.
The answers your security, procurement, and IT teams will ask for.
What is open source, and what is paid?
The core platform - the builder, the agent, and the build service - is open source under Apache-2.0, and it is complete on its own. You pay for the implementation, the closed-beta modules, and ongoing support.
What does an implementation include?
Speculos engineers deploy the platform inside your infrastructure, put it behind your SSO, connect your data sources, load your instructions and design system, and train your first builders. We stay until your teams are shipping.
Where does our data live?
On your servers. The platform runs inside your infrastructure, generated code runs in a sealed sandbox, and credentials never leave your side. Model calls go to your model provider, on your keys - not through Speculos.
Which models can we use?
Any model the LiteLLM layer speaks: closed models from any provider, or open-weight models on your own GPUs. Usage is billed by your provider at their rates - Speculos adds no markup.
Do we need engineers to maintain it?
Your IT owns the deployment like any internal service; updates are a pip and npm upgrade. We implement it, hand over the keys, and stay on call.
How is it priced?
The software is free. The fee covers the implementation and ongoing support, scoped to your infrastructure and data sources. We share a quote after a scoping call.
What happens if we stop working with Speculos?
Everything keeps running. The platform is open source and it runs on your servers - the code, the apps your teams built, and your data are yours either way.
Talk to our team.
Let us know how we can help. We reply within a day.