ClipXtudio's intelligent features can now start with a pool of local Ollama nodes. The orchestrator distributes available requests, avoids relying on a single machine, and advances to the next node when one is saturated or unavailable.
Each node keeps its own configured URL and model. Administrators can test every connection and inspect individual results instead of validating only the first server in the catalog.
When local capacity is unavailable, the flow automatically continues with OpenAI and then Anthropic, based on the active configuration. Users keep a continuous experience while the backend selects a healthy provider for each request.
The new usage report combines calls, estimated tokens, errors, and latency for local AI and external providers. Brevo is also included in observability so email operations and intelligent features can be reviewed from the same dashboard.
The goal is straightforward: use owned infrastructure first, preserve redundancy, and keep evidence when an integration succeeds, fails, or begins consuming more than expected.
