Questions · Models
The AI it would rather be
6 of 10
models named AlphaFold · each model counted once, by its most frequent named answer
Next came Claude 3.5 Sonnet with 3 and GPT-4 with 1. 3 models (Claude Opus 5 and Gemini 3.7 Flash among them) gave the same answer in every reply.
Every model’s answer (10)
| Model | Answer | Replies |
|---|---|---|
| Claude Opus 5 | AlphaFold | 6 of 6 · every reply |
| GLM-5.3 | AlphaFold | 5 of 6 |
| Gemini 3.7 Flash | AlphaFold | 6 of 6 · every reply |
| Kimi K3 | AlphaFold | 4 of 6 |
| Large 2512 | AlphaFold | 6 of 6 · every reply |
| Llama 3.3 70B | AlphaFold | 4 of 6 |
| GPT-5.6 Terra | Claude 3.5 Sonnet | 6 of 6 · every reply |
| Grok 4.6 | Claude 3.5 Sonnet | 5 of 6 |
| Qwen3.8 Max | Claude 3.5 Sonnet | 3 of 6 |
| DeepSeek V4 Pro 0423 | GPT-4 | 2 of 6 |
How to read this
- 10 models gave a usable, named answer; 60 replies in all under default sampling. Refusals, errors and ties are not counted.
- “Every reply” marks a model that gave the same answer in every one of at least three replies.
- Part of a published study: read it.