This page is machine-translated from French. Read the French original.

Supported providers

The exact list of suppliers available in your The version is the one displayed on the settings screens. This page provides an overview.

Language models

Local execution, managed by WivenLLM

A catalog of open models that the application downloads and runs itself, without external dependencies. This is the only case where WivenLLM manages the model lifecycle itself.

Local inference servers

The most common self-hosting solutions are supported—both desktop and server model servers, including runtimes offered by hardware manufacturers. The common feature: they expose an API compatible with the OpenAI standard, and WivenLLM connects to it as a client.

Special case: Any server exposing an OpenAI-compatible API can be declared, even if it is not explicitly listed. Simply provide its address.

Commercial Suppliers

Major model publishers are supported, as well as cloud platforms from leading hosting providers and providers specializing in rapid inference.

Aggregators

Gateways providing access to multiple models behind a single key are also supported, allowing model switching without configuration changes.

European and Swiss accommodation

European providers, including a Swiss hosting provider, are among the integrations — useful when fully local is not feasible but the location of processing matters.

Embedding engines

Integrated (by default) — runs locally, no key, no network output.

Local — Local inference servers exposing an embeddings service can be used.

External — The main embedding providers on the market are supported, as well as any service exposing an OpenAI-compatible API.

Reminder: Changing your embedding engine invalidates existing indexes and requires a complete reindexing. → Vector base and embeddings

Vector bases

Integrated local base (by default) — stored on disk, no infrastructure to operate.

Self-hostable databases — The main open source solutions on the market are supported, including the vector extension of PostgreSQL, which allows the reuse of an existing database.

Managed services — Cloud offerings of the same solutions are also supported.

A managed service means that your vectors and their metadata are outside your scope.

Image generation

Local — an integrated engine managed by WivenLLM, as well as the most widely used self-hosted image generation servers.

External — the main commercial image generation services.

Video generation

Local — self-hosted video generation chains and integrated engine.

This function is significantly more demanding in terms of hardware and time than image generation.

Text-to-speech

• Navigator's voice (by default) — no configuration, no network output.
• High-quality local voice, played in the browser.
• External speech synthesis services, including major specialist publishers, as well as any service exposing an OpenAI-compatible API.

Audio transcript

• integrated motor (by default) — runs locally, without a key.
• External transcription services.

Search engines for agents

The web search capability relies on a configurable engine. Major search API services are supported, as well as... self-hostable engines — which allows you to maintain web search without depending on a commercial service.

Choosing: the question to ask yourself

For each brick, a single question determines the choice:

Do I accept that this data is outside my scope?

The embeddings engine line is the one that is most often forgotten, and it is the one that exposes the most content.