AI tools I use (and why)

Which AI tools I use and why. Last update: 25-04-18

• I use different models for different tasks.
  • I use GPT 4.1 Mini as the default LLM (cheap, fast, good).
  • I use Claude Sonnet 3.7 (best, but expensive) and Deepseek v3 (cheap, very good) for coding (via aider)
  • I use Gemini Flash 2 for OCR
• I use aider for pair programming with LLMs
• I use the llm Python tool to test and use LLM APIs and for chat-type LLM interaction with all sort of LLMs(Link).
  • (+) It supports all models that I use and test frequently.
  • (+) It logs previous prompts.
  • (+) It also has support for embeddings and topic clustering.
• I use Google's AI Studio as writing companion, esp. with Gemini 2.5 Pro
• I use the paid Perplexity Pro that features latest models
  • (+) I like that I can switch between models in a chat format.
  • (+) I like how it integrates web search.
  • (-) The results are worse imho than 6 months ago, e.g. worse Reddit summaries
• I have no established tool for evaluating LLMs programmatically and systematically yet, I'm considering PydanticAI Evals, ChainForge, AlignEval and PromptFoo at the moment