100% open source · inspect · modify · self-host

Keep the
experience.
Choose the intelligence.

Claude and ChatGPT pair a great product with their own models. Liberde unbundles the two: one polished experience where Claude, GPT, Kimi, Qwen, Mistral, Llama, and other models work together.

An abstract bird leaving a closed structure, representing freedom from model lock-in.
No provider lock-in
Your experience stays.
The model is your choice.

Design. Converse. Create.

Use the experience you love without giving control to one provider. Move between models, keep your work, and understand the cost, data use, and environmental impact. Freely. Responsibly. Transparently.

Liberde / Conversation Control 6 models available

Change models, not conversations.

Use Claude for one response, GPT for the next, and an open or local model when privacy, cost, or control matters. The conversation, context, files, and tools stay in place.

Choose the model for the next response
Impact measured live

One conversation, fully under your control

01 Next response Claude Your choice
02 Shared conversation Context, files, memory, and tools stay available Portable
03 Estimated environmental impact 0.42 Wh energy · 0.18 gCO₂e this conversation Visible
04 Usage & budget 2,418 tokens · 64% below your limit Controlled
Response time 842ms Response cost $0.038

Your AI use, under your control.

Model choice is only the start. Liberde also shows the tokens, cost, data access, tool use, and estimated environmental impact behind each response.

01 / UNBUNDLED

Keep the UX. Change the model.

Get one polished experience without tying your conversations, context, and workflows to a single model provider.

02 / TOGETHER

Use many models in one conversation.

Move between closed, open, specialist, and local models. Each one can pick up where the last one stopped.

03 / TRANSPARENT

See the things that matter.

Check tokens, cost, context, permissions, tool calls, data access, and estimated environmental impact for every response.

Nothing hidden

Pick the model.
Keep the context.
Know the impact.

Open, closed, specialist, and local models available 100s
Conversations lost when you change models 0
Token usage and cost visible for each response Always
Estimated energy use and carbon impact Tracked
Source code available to inspect and control 100% open
“The experience you love should not decide which intelligence you can use.
— Why we built Liberde
An abstract bird leaving a closed structure, representing freedom from model lock-in. One UX
every model
your control
L

Free your AI experience

One experience.
Every model.

Keep the UX, conversations, context, files, and tools you rely on. Choose the best model for each response—with clear cost, token use, data access, and environmental impact.