AI tooling is currently moving so fast it is imposible to keep up with all tools and all developments.
This repo is an attempt to document specific time-bound experiments related to AI.
- The experiments are time bound
- The experiments are documented
- The experiments are repeatable
I expect repeated experiments over time to demonstrate different results. Hopefully better results.
- https://chatgpt.com/
- https://lambda.chat/
- https://duck.ai
- https://grok.com/
- https://www.perplexity.ai/
- https://deepai.org/
- https://chat.mistral.ai/chat
- https://claude.ai
- https://chat.z.ai/
- https://www.kimi.com/
- https://chat.qwen.ai/
- https://aistudio.google.com/prompts/new_chat
- https://gemini.google.com/app
- https://poe.com/
- https://chat.deepseek.com/
- https://openrouter.ai/ there are free models, can be very hit or miss if they work or not
- https://www.phind.com/
- https://lmarena.ai/ (set to one model at a time https://lmarena.ai/?mode=direct)
Remember not to put any confidential information in online cloude based tooling. Particularly on free plans, but probably not at all, and remember to check settings in the cloud tools to make sure that you are not allowing someone to review the chats and responses to help train the AI further.
- ollama.com
- https://www.nomic.ai/gpt4all
By running the models locally we can use confidentially information without worrying too much.
Locally I've been using:
- "qwen3:14b" with Ollama for general chat
- "CodeLlama" with Ollama for code generation
With Gpt4All for general chat:
- Reasoning 1
- Llama 3b instruct