Image input

Jump to bottom

Takashi Iwamoto edited this page Feb 24, 2025 · 1 revision

Usage

Set a model that supports image input in the environment variable LITELLM_MODEL.
Set true in the environment variable IMAGE_FILE_ACCESS_ENABLED.
Run Collmbo.
Upload your image(s) and send a message in Slack.

$ cat env
SLACK_APP_TOKEN=xapp-1-...
SLACK_BOT_TOKEN=xoxb-...
OPENAI_API_KEY=sk-...
LITELLM_MODEL=gpt-4o
IMAGE_FILE_ACCESS_ENABLED=true

$ docker run -it --env-file ./env ghcr.io/iwamot/collmbo:latest-slim