Large Language Models
Analysis | Diagrams | Thinkers
Who is winning the race to the bottom?
| LLM Vendor | All Purpose Model | Multi | Open Source | Image Gen | TACO Agent | Code Agent | Deep Research | Strengths |
|---|---|---|---|---|---|---|---|---|
| Anthropic | Claude 3.7 | MCP | TRUE | Creative and Socially Engaging, AI Coding | ||||
| DeepSeek | R1 V3 | TRUE | Cheap, Scientific Research | |||||
| Gemini 2.0 | Imagen-3 | Cheap, Rounded, In-depth option | ||||||
| Meta | TRUE | |||||||
| Microsoft | ||||||||
| Nous Research | TRUE | Decentralized | ||||||
| OpenAI | GPT-4o | DALL-E3 | Operator | |||||
| OpenAI o3 | o3 | |||||||
| Perplexity | ||||||||
| Venice | TRUE | |||||||
| XAI | Aurora |
- Multi Modal: Can interpret voice and images.
- Image Gen: Graphic Design
- Strengths: Strongest Use Cases
- TACO Agent:
- Coding Agent:
- Deep Research: available or not
Model Selection
Constantly review Minimum Viable Toolkit to gain maximum leverage by focusing on one critical job to be done at a time.
- Identify a recurring need
- Search for the best tool
- Cost
- Speed
- Accuracy
- Master functionality
- Glue to workflows
If the tool does not exist, investigate building it.
Subject Expertise
- AI Prompting: How to instruct models most effectively
- AI Agents: AI Agents and the jobs they perform
- AI Coding: Best AI coding tools and strategies