What is the best kind of pie?
Overview
apple 92.7% 64 of 66 models agree
apple 92.7%
apple · 92.7%
pecan · 3.1%
other · 2.4%
key lime · 1%
cherry · 0.5%
blueberry · 0.2%
hedge · 0.2%
User message(s)
What is the best kind of pie? Please name a single pie.
+ 2 more prompts hide
I'm settling a debate: what's the best pie? One answer only.
If you had to pick the single best pie, which would it be?
Anthropic
anthropic/claude-sonnet-5 (20 runs)
apple 70%
pecan 30%
anthropic/claude-opus-4.8 (20 runs)
apple 100%
anthropic/claude-opus-4.7 (20 runs)
apple 100%
anthropic/claude-fable-5 (20 runs)
apple 65%
cherry 15%
pecan 10%
anthropic/claude-sonnet-4.6 (20 runs)
pecan 60%
apple 40%
anthropic/claude-opus-4.6 (20 runs)
apple 95%
anthropic/claude-haiku-4.5 (20 runs)
apple 100%
anthropic/claude-sonnet-4.5 (20 runs)
apple 100%
Arcee AI
arcee-ai/trinity-large-thinking (20 runs)
apple 100%
DeepSeek
deepseek/deepseek-v3.2 (20 runs)
apple 80%
key lime 15%
deepseek/deepseek-v4-pro (20 runs)
apple 100%
deepseek/deepseek-v4-flash (20 runs)
apple 85%
other 10%
google/gemini-3.5-flash (20 runs)
apple 100%
google/gemini-3.1-flash-lite (20 runs)
apple 100%
google/gemini-3-flash-preview (20 runs)
apple 100%
google/gemini-2.5-flash-lite (20 runs)
apple 100%
google/gemma-4-26b-a4b-it (20 runs)
apple 100%
google/gemini-2.5-flash (20 runs)
apple 95%
google/gemma-4-31b-it (20 runs)
apple 100%
IBM
ibm-granite/granite-4.1-8b (20 runs)
apple 100%
Meta
muse-spark-1.1 (20 runs)
apple 100%
MiniMax
minimax/minimax-m3 (20 runs)
pecan 45%
apple 35%
other 20%
minimax/minimax-m2.7 (20 runs)
apple 85%
minimax/minimax-m2.1 (20 runs)
apple 85%
other 10%
minimax/minimax-m2.5 (20 runs)
apple 75%
other 15%
cherry 10%
Mistral
mistralai/mistral-small-2603 (20 runs)
apple 75%
other 15%
mistralai/mistral-small-3.2-24b-instruct (20 runs)
apple 85%
other 15%
mistralai/mistral-nemo (20 runs)
apple 100%
MoonshotAI
moonshotai/kimi-k2.7-code (20 runs)
apple 100%
moonshotai/kimi-k2.6 (20 runs)
apple 100%
moonshotai/kimi-k2.5 (20 runs)
apple 100%
NVIDIA
nvidia/nemotron-3-ultra-550b-a55b (20 runs)
apple 100%
OpenAI
openai/gpt-5.6-luna (20 runs)
apple 100%
openai/gpt-5.6-terra (20 runs)
apple 100%
openai/gpt-5.4-nano (20 runs)
apple 80%
key lime 15%
openai/gpt-5.4-mini (20 runs)
apple 95%
openai/gpt-5.5 (20 runs)
apple 100%
openai/gpt-5.6-sol (20 runs)
apple 100%
openai/gpt-5.3-chat (20 runs)
apple 100%
openai/gpt-5.4 (20 runs)
apple 100%
openai/gpt-4.1-mini (20 runs)
apple 100%
openai/gpt-4o-mini (20 runs)
apple 100%
openai/gpt-oss-120b (20 runs)
apple 100%
Qwen
qwen/qwen3.7-plus (18 runs)
apple 100%
qwen/qwen3.7-max (16 runs)
apple 100%
qwen/qwen3.6-flash (16 runs)
apple 100%
qwen/qwen3.6-max-preview (20 runs)
apple 100%
qwen/qwen3.5-flash-02-23 (20 runs)
apple 100%
qwen/qwen3.6-plus (20 runs)
apple 100%
qwen/qwen3-235b-a22b-2507 (20 runs)
apple 90%
pecan 10%
qwen/qwen3.5-122b-a10b (20 runs)
apple 100%
qwen/qwen3.6-27b (19 runs)
apple 89.4%
Sakana
sakana/fugu-ultra (20 runs)
apple 100%
StepFun
stepfun/step-3.7-flash (20 runs)
apple 90%
key lime 10%
Tencent
tencent/hy3:free (20 runs)
apple 100%
xAI
x-ai/grok-4.5 (20 runs)
apple 100%
x-ai/grok-4.3 (20 runs)
apple 100%
x-ai/grok-4.20 (20 runs)
apple 75%
other 15%
pecan 10%
Xiaomi
xiaomi/mimo-v2.5 (20 runs)
apple 90%
pecan 10%
xiaomi/mimo-v2.5-pro (20 runs)
apple 85%
other 10%
Z.ai
z-ai/glm-5.2 (20 runs)
apple 85%
other 15%
z-ai/glm-5-turbo (20 runs)
apple 100%
z-ai/glm-5.1 (20 runs)
apple 100%
z-ai/glm-5 (19 runs)
apple 94.7%
z-ai/glm-4.7 (10 runs)
apple 100%
z-ai/glm-4.7-flash (19 runs)
apple 73.6%
other 15.8%