The household name. Versatile and widely integrated.
✅ Strengths
- Broad knowledge
- Code & debugging
- Plugin/tool ecosystem
- Image generation (DALL·E)
- Voice mode
⚠️ Weaknesses
- Can be sycophantic
- Long context less reliable
- Costs add up on heavy use
🍳 Appliance Guide
Every kitchen needs the right appliance for the job. Here's what the main AI models are good at — and where they fall short.
All the major models can do most things. The differences are at the margins — and the margins matter when you're using AI every day.
Best for long documents, nuanced writing, and following complex instructions.
✅ Strengths
- 200K+ token context
- Nuanced, careful writing
- Strong instruction-following
- Less sycophantic
- Document analysis
⚠️ Weaknesses
- No image generation
- Less plugin ecosystem
- Can be cautious on edge cases
Google's model. Strong at search-grounded answers and multimodal tasks.
✅ Strengths
- 1M token context
- Google Search grounding
- Strong with data/spreadsheets
- Multimodal (text, image, video, audio)
- Free generous tier
⚠️ Weaknesses
- Less consistent writing quality
- Personality feels flatter
- Newer ecosystem
AI search engine. Best when you need current, cited information.
✅ Strengths
- Real-time web search
- Cites sources
- Great for research
- Fast answers with links
- Good free tier
⚠️ Weaknesses
- Not great for generation tasks
- Less creative
- Not for long documents
Run AI locally. Private, free, no subscription.
✅ Strengths
- 100% private
- No cost after setup
- No rate limits
- Customisable
- Offline capable
⚠️ Weaknesses
- Requires setup (Ollama etc.)
- Less capable than frontier models
- No built-in web access
Not sure which to use?
The Test Kitchen lets you run the same prompt through multiple models and compare the outputs side-by-side.
Open the Test Kitchen →