I have a vastly updated version of my Image Description Toolkit (IDT) available. The 3.0 release builds on the command line workflow tools I released earlier with a guided workflow creation system, analysis tools for combining descriptions, a range of performance statistics on your descriptions and tools to review content from various AI models and prompts.
The IDT also includes several other tools for working with image descriptions. A results viewer allows you to browse results, monitor workflows as they happen, copy descriptions and images.
If you want to focus on detailed descriptions for individual images, the Image Describer tool, allows you to load a directory of images, and easily request descriptions for individual images and do so using multiple providers and prompts and save all your work for future use.
A Prompt Editor allows you to create and edit the prompts used to guide AI descriptions for all the tools in the IDT and configure defaults.
In addition to Ollama, which powered earlier versions of the IDT, the IDT now includes support for multiple AI models from both OpenAI and Anthropic if you provide your own API key.
A comprehensive User Guide is available for the various tools in the IDT.
Whether you want detailed descriptions for one or one thousand images, the IDT has a tool for you. Grab IDT3.zip, extract the files to your computer and run install_idt. Be sure to install Ollama and or get API keys for OpenAI or Anthropic and configure your AI models. Then open a command prompt and run:
idt guideme to generate your first descriptions.
Onedrive link doesn’t work
Could you try this link:
https://1drv.ms/f/c/a7b1bd807b044bbc/EpUH56shTt5Mo6kOv7mTbnUBuExso61DMl0aQM2nUPxWsQ?e=TkXDrR
OK, now it works. But I cannot use this tool, because I have two problems. First is that even if I’m providing OpenAI working apikey, I cannot chose OpenAI from the providers list. I am experiencing the second issue when I’m choosing Ollama Cloud model qwen3-vl:235b-cloud, I can send images in the Ollama gui and get descriptions, but when I’m choosing it from Image Describer, I am receiving following error:
⚠️ Ollama Cloud model ‘qwen3-vl:235b-cloud’ doesn’t support vision capabilities yet.
💡 Try these local vision models instead:
• llava:latest (7B parameters)
• llava-llama3:latest (8B parameters)
• bakllava:latest (7B parameters)
• moondream:latest (1.8B parameters)
Cloud models are excellent for text-only tasks but vision support is coming soon!
Do you have any idea on how to fix that?
How are you getting to the cloud model with the tool. The error you showed is exactly what I get if I try and use the cloud model from Ollama directly and is one of the reasons I added it to my tool.
Same thing with OpenAI. Are you for example using idt guideme or a different way of trying this?
Hello. I just thought, that if I’ll declare my openai api key in the prompt editor, that will be enough for using GPT. But OK, this problem was solved, but I have another issue. In prompt editor I have a list of 3 prompts, simple, artistic and my custom prompt. But when I’ll type idt prompt-list I’m receiving only a list of 7 predefined prompts. The last line of this list contains following information: “Config file: C:\Users\Paulinux\AppData\Local\Temp\_MEI144522\scripts\image_describer_config.json”. Unfortunately, this location isn’t available. When I’m declaring the prompt style to use with idt guideme –prompt-style mypromptname, I’m receiving the description that is telling in metadata that the prompt style used was my custom, but I know that style actually used is default, narrative. I know that because my prompt is in other language so the model is answering properly. And I could test my prompt style inside Image Describer. What should I do in this situation? With regards, Paulina
Are you saying that if you use the Prompt Editor and make a new prompt and save the updated prompt file to the defualt location, the new prompt works in the Image Describer app but not when you type idt prompt-list? I’ll need to investigate that one a bit more.
Yes, on my second computer this issue persists
Hi,
I hope to have a fix for you on all this configuration area in the next day or so. I’m testing some changes to ensure specifying custom configuration files works end-to-end.
Hi,
It sounds like you figured out how to create custom config files. I’ve made sizable updates to all of this and believe it should word the way you want now.
1. Create your custom prompt file however you know to do it.
2. when running idt commands add the option –config–id
There is also a new option on the file menu to load your own prompt file. If you know how to create those, just load that and in those files you can set your own defaults.
For example, this command loads my custom prompts file and my custom video processing file. I keep these in a directory named kelly off the scripts folder where I’ve installed idt.
idt workflow –config-id scripts\kelly\descriptionskelly.json –config-video scripts\kelly\videokelly.json \\ford\home\photos\mobilebackup\iphone\2023\04
There are links to full documentation on this all at the end of this reply.
I’m not sure where you were downloading from but if you go to http://www.theideaplace.net/projects you’ll find links. You want the 3.5 beta link.
You can also download directly from:
https://1drv.ms/u/c/a7b1bd807b044bbc/EZZVbOsOtDpAs0BDC17X6gYBNgaAAKvWL0juAQpe6vTW3Q?e=SP532Z
References:
Main Documentation:
– USER_GUIDE.md: https://github.com/kellylford/Image-Description-Toolkit/blob/feature/explicit-config-arguments/docs/USER_GUIDE.md
– CLI_REFERENCE.md: https://github.com/kellylford/Image-Description-Toolkit/blob/feature/explicit-config-arguments/docs/CLI_REFERENCE.md
– WHATS_NEW_v3.5.0-beta.md: https://github.com/kellylford/Image-Description-Toolkit/blob/feature/explicit-config-arguments/docs/WHATS_NEW_v3.5.0-beta.md
– README.md (docs index): https://github.com/kellylford/Image-Description-Toolkit/blob/feature/explicit-config-arguments/docs/README.md
Configuration & Customization:
– CONFIGURATION_GUIDE.md: https://github.com/kellylford/Image-Description-Toolkit/blob/feature/explicit-config-arguments/docs/CONFIGURATION_GUIDE.md
– PROMPT_WRITING_GUIDE.md: https://github.com/kellylford/Image-Description-Toolkit/blob/feature/explicit-config-arguments/docs/PROMPT_WRITING_GUIDE.md
– WEB_DOWNLOAD_GUIDE.md: https://github.com/kellylford/Image-Description-Toolkit/blob/feature/explicit-config-arguments/docs/WEB_DOWNLOAD_GUIDE.md
Thank you very much. I’ll test it soon.
OK, I encountered the next error. Now, when editing the default config file, the prompt list in the IDT cli is identical to prompts showed in the prompt editor. That’s very nice and thanks for this repair! But when I’m running workflow using this command:
idt workflow c:\biedra –provider openai –model gpt-4o-mini –output-dir Descriptions –api-key-file C:\Users\Paulinux\openai.txt –prompt-style Biedra –no-metadata
I am receiving blank output:
INFO – Total files processed: 0 – (2025-11-04 13:19:51,901)
despite I have files in this folder, from 0.jpg to 19.jpg
What I’m doing wrong?
First off, if you are able to talk by voice in some fashion, I’d be happy to try and sort this out with you more directly. Or even email, write me directly at kelly@theideaplace.net.
Second, in the directory you specified for your output, there should be a folder starting with wf and then some form of the model provider and such you used. Inside of that there is a directory called logs. Can you read those log files and see if they show any errors. There is an overall workflow log and then logs for each step in the process.
My initial hunch would be that something went wrong with the OpenAI key but it is hard to know for sure.
Also at the same level as logs, there is a folder called descriptions. Does that have any files in it? If everything works there will be a text file. for some failures, if things go wrong, there will be a failure file.
Are you creating these commands with the idt guideme approach or crafting the fuoll command yourself.
If you haven’t tried the guide me option, please do once and see if it works to create your run command.
Hi,
I also put one more update that you can download that may solve the issue you are describing but I’d still storngly suggest checking the logs and trying the guideme option as well. You can download this update form the same places you’ve obtained earlier versions. You still want the 3.5 beta and can install over your existing version.
Thank you. I’ll test it soon
Wow, it works now. Thank you very much for your support and guidance!