How do I find out which are the best LLMs to use for different tasks?
LLM
Sources of rankings and info on LLMs
Up to date sources for LLM ranking:
- Use LLM Leaderboard 2025 from Vellum for main benchmarks.
- Trending LLM Rankings from OpenRouter to see which are trending.
- OpenRouter also has very nice descriptions of all the models.
- LLM Leaderboard 2025 - Compare LLMs has a nice graph.
- Models from OpenRouter is great for filtering on price etc.
Currently:
- DeepSeek R1 is a good option for research.
- Gemini 2.5 Pro looks very good for search queries.
- Claude 3.7-Sonnet is still one of the best for coding, but OpenAI GPT-4.1 Mini looks like a new strong contender and Google: Gemini 2.0 Flash also looks good.