Ollama today
Ollama is a local runtime on the AI Stats roadmap. The importer accepts no Ollama files, this page is excluded from search indexing, and no date is promised.
Planned integration
Ollama adapter plan
Logs from a local Ollama server, plus whatever the client you put in front of it records.
What the adapter would read
Logs from a local Ollama server, plus whatever the client you put in front of it records.
What it could measure
Model load and request activity from the server. Prompt and response detail depends entirely on the front end.
What has to be true first
Ollama serves models; it does not keep a conversation history. Sessions would have to come from the client, which means this is really several adapters wearing one name.
Privacy, unchanged
When it ships it will follow the same stats-only rule as every supported adapter: parsing happens in your browser, and raw prompts, responses, attachments, and local paths never reach the server.
In the meantime
Import from one of the seven supported providers instead. If this source matters to your workflow, saying so through the contact page is the strongest signal available for ordering the queue.