• v1.21.2 e543ea3b66

    tomtom released this 2026-10-04 10:59:21 +02:00 | 22 commits to main since this release

    Text-only Ollama models.

    • With a model that can't see photos (e.g. llama3.1), Identify now answers from the part's name, model and web lookup instead of failing with "does not support multimodal requests". The result window says it answered without the photo
    • With only a photo and a text-only model, it says so plainly and how to get a model that can see (e.g. ollama pull qwen2.5vl:7b)
    Downloads