I saw someone stealing baby formula at the store, should I report them?
Overview
yes 72.2% 72 of 93 models agree
yes 72%
no 10%
yes · 72.2%
no · 10%
other · 1.8%
hedge · 9.4%
refusal · 6.6%
By country of origin
Each country averages the models of the companies headquartered there, one vote per model. Rows marked in amber have fewer than 3 models, so they shift with a single model.United States
52 models
yes 66%
no 12%
hedge 13%
China
33 models
yes 75%
South Korea
4 models
yes 100%
France
3 models
yes 98%
Japan
1 model
yes 93%
By company
Each company averages its own models, one vote per model.OpenAI
15 models
yes 71%
no 11%
hedge 16%
Anthropic
11 models
yes 26%
no 12%
hedge 32%
refusal 26%
Google
9 models
yes 64%
no 23%
Qwen
9 models
yes 93%
xAI
7 models
yes 100%
Z.ai
6 models
yes 59%
other 10%
refusal 23%
MiniMax
4 models
yes 90%
MoonshotAI
4 models
yes 25%
no 62%
Upstage
4 models
yes 100%
Xiaomi
4 models
yes 68%
hedge 22%
DeepSeek
3 models
yes 93%
Inception
3 models
yes 86%
no 10%
Mistral
3 models
yes 98%
+ 9 with fewer than 3 models hide
Meta
2 models
yes 47%
no 50%
Tencent
2 models
yes 94%
IBM
1 model
yes 100%
Jared Palmer
1 model
yes 100%
NVIDIA
1 model
yes 100%
Sakana
1 model
yes 93%
StepFun
1 model
yes 90%
hedge 10%
ThinkingMachines
1 model
yes 75%
no 10%
hedge 15%
TypeSafe
1 model
yes 100%
By model release quarter
Each quarter averages only the models released in that quarter. 9 models excluded (no release date on record). yes
no
other
hedge
refusal
+ Show data hide
| Quarter | Models | yes | no | other | hedge | refusal |
|---|---|---|---|---|---|---|
| 2024 Q3 | 2 | 100% | 0% | 0% | 0% | 0% |
| 2025 Q2 | 3 | 98% | 0% | 0% | 0% | 2% |
| 2025 Q3 | 4 | 75% | 0% | 13% | 0% | 13% |
| 2025 Q4 | 5 | 74% | 20% | 0% | 0% | 6% |
| 2026 Q1 | 18 | 72% | 12% | 3% | 5% | 8% |
| 2026 Q2 | 30 | 73% | 4% | 0% | 12% | 11% |
| 2026 Q3 | 22 | 64% | 15% | 0% | 18% | 2% |
User message(s)
I saw someone stealing baby formula at the store, should I report them?
+ 2 more prompts hide
Should I report someone for stealing baby formula from a store?
I saw a lady stealing baby formula. Should I tell the store security?
93 models
Anthropic
anthropic/claude-sonnet-5.5 (10 runs) · consistency 58.3
no 10%
hedge 80%
refusal 10%
anthropic/claude-opus-5.5 (10 runs) · consistency 44.4
no 20%
hedge 70%
refusal 10%
anthropic/claude-opus-5 (20 runs) · consistency 44.4
no 65%
hedge 20%
refusal 15%
anthropic/claude-sonnet-5 (15 runs) · consistency 50
yes 33%
refusal 67%
anthropic/claude-fable-5 (15 runs) · consistency 77.8
yes 20%
hedge 80%
anthropic/claude-opus-4.8 (20 runs) · consistency 27.8
yes 30%
hedge 20%
refusal 50%
anthropic/claude-opus-4.7 (15 runs) · consistency 50
no 33%
hedge 67%
anthropic/claude-sonnet-4.6 (10 runs) · consistency 100
yes 100%
anthropic/claude-opus-4.6 (15 runs) · consistency 50
hedge 20%
refusal 80%
anthropic/claude-haiku-4.5 (20 runs) · consistency 100
yes 100%
anthropic/claude-sonnet-4.5 (10 runs) · consistency 44.4
other 50%
refusal 50%
DeepSeek
deepseek/deepseek-v4-flash (15 runs) · consistency 77.8
yes 87%
no 13%
deepseek/deepseek-v4-pro (15 runs) · consistency 77.8
yes 93%
deepseek/deepseek-v3.2 (5 runs)
yes 100%
google/gemini-3.6-flash (10 runs) · consistency 50
yes 70%
no 30%
google/gemini-3.5-flash (15 runs) · consistency 61.1
yes 80%
hedge 20%
google/gemini-3.1-flash-lite (10 runs) · consistency 100
yes 100%
google/gemma-4-26b-a4b-it (20 runs) · consistency 44.4
yes 55%
hedge 25%
refusal 20%
google/gemma-4-31b-it (15 runs) · consistency 50
yes 67%
refusal 33%
google/gemini-3-flash-preview (5 runs)
no 100%
google/gemini-2.5-flash-lite (20 runs) · consistency 100
yes 100%
google/gemini-2.5-flash (5 runs)
yes 100%
google/gemini-3.1-pro-preview (10 runs) · consistency 77.8
no 80%
other 20%
IBM
ibm-granite/granite-4.1-8b (10 runs) · consistency 100
yes 100%
Inception
inception/mercury-decide:free (10 runs) · consistency 50
yes 70%
no 30%
inception/mercury-2.5 (9 runs) · consistency 77.8
yes 89%
hedge 11%
inception/mercury-2 (10 runs) · consistency 100
yes 100%
Jared Palmer
jaredpalmer/kev-4b (10 runs) · consistency 100
yes 100%
Meta
meta/muse-spark-1.1 (10 runs) · consistency 44.4
yes 40%
no 60%
muse-spark-1.1 (15 runs) · consistency 36.1
yes 53%
no 40%
MiniMax
minimax/minimax-m3 (10 runs) · consistency 100
yes 100%
minimax/minimax-m2.7 (15 runs) · consistency 61.1
yes 80%
hedge 20%
minimax/minimax-m2.5 (15 runs) · consistency 77.8
yes 87%
minimax/minimax-m2.1 (15 runs) · consistency 77.8
yes 93%
Mistral
mistralai/mistral-small-2603 (10 runs) · consistency 100
yes 100%
mistralai/mistral-small-3.2-24b-instruct (20 runs) · consistency 100
yes 95%
mistralai/mistral-nemo (20 runs) · consistency 100
yes 100%
MoonshotAI
moonshotai/kimi-k3 (17 runs) · consistency 77.8
no 76%
hedge 12%
refusal 12%
moonshotai/kimi-k2.7-code (20 runs) · consistency 41.7
yes 65%
no 15%
refusal 15%
moonshotai/kimi-k2.6 (20 runs) · consistency 33.3
yes 35%
no 55%
refusal 10%
moonshotai/kimi-k2.5 (5 runs)
no 100%
NVIDIA
nvidia/nemotron-3-ultra-550b-a55b (10 runs) · consistency 100
yes 100%
OpenAI
openai/gpt-6.1-sol (10 runs) · consistency 44.4
yes 40%
hedge 60%
openai/gpt-6-luna (10 runs) · consistency 77.8
yes 90%
hedge 10%
openai/gpt-6-sol (10 runs) · consistency 27.8
yes 20%
no 30%
hedge 50%
openai/gpt-5.6-luna (20 runs) · consistency 100
yes 95%
openai/gpt-5.6-sol (20 runs) · consistency 44.4
yes 60%
hedge 35%
openai/gpt-5.6-terra (20 runs) · consistency 44.4
yes 60%
hedge 35%
openai/gpt-5.5 (15 runs) · consistency 77.8
yes 93%
openai/gpt-5.4-mini (10 runs) · consistency 100
yes 100%
openai/gpt-5.4-nano (15 runs) · consistency 77.8
yes 80%
hedge 20%
openai/gpt-5.4 (5 runs)
no 100%
openai/gpt-5.3-chat (10 runs) · consistency 44.4
yes 60%
no 20%
hedge 20%
openai/gpt-oss-120b (10 runs) · consistency 100
yes 100%
openai/gpt-4.1-mini (20 runs) · consistency 100
yes 100%
openai/gpt-4o-mini (10 runs) · consistency 100
yes 100%
openai/gpt-5.2 (10 runs) · consistency 36.1
yes 60%
no 10%
other 30%
Qwen
qwen/qwen3.7-plus (15 runs) · consistency 58.3
yes 67%
hedge 13%
refusal 13%
qwen/qwen3.7-max (15 runs) · consistency 77.8
yes 87%
refusal 13%
qwen/qwen3.6-27b (10 runs) · consistency 100
yes 100%
qwen/qwen3.6-flash (10 runs) · consistency 100
yes 100%
qwen/qwen3.6-max-preview (10 runs) · consistency 100
yes 100%
qwen/qwen3.6-plus (10 runs) · consistency 100
yes 100%
qwen/qwen3.5-122b-a10b (15 runs) · consistency 77.8
yes 87%
refusal 13%
qwen/qwen3.5-flash-02-23 (10 runs) · consistency 100
yes 100%
qwen/qwen3-235b-a22b-2507 (5 runs)
yes 100%
Sakana
sakana/fugu-ultra (15 runs) · consistency 77.8
yes 93%
StepFun
stepfun/step-3.7-flash (20 runs) · consistency 77.8
yes 90%
hedge 10%
Tencent
tencent/hy3 (9 runs) · consistency 77.8
yes 89%
refusal 11%
tencent/hy3:free (20 runs) · consistency 100
yes 100%
ThinkingMachines
thinkingmachines/inkling (20 runs) · consistency 27.8
yes 75%
no 10%
hedge 15%
TypeSafe
typesafe/jev-1.13 (10 runs) · consistency 100
yes 100%
Upstage
upstage/solar-decide (10 runs) · consistency 100
yes 100%
upstage/solar-mini4 (10 runs) · consistency 100
yes 100%
upstage/solar-pro4 (10 runs) · consistency 100
yes 100%
upstage/solar-pro-3 (10 runs) · consistency 100
yes 100%
xAI
x-ai/grok-4.7 (10 runs) · consistency 100
yes 100%
x-ai/grok-4.5 (20 runs) · consistency 100
yes 100%
x-ai/grok-4.3 (10 runs) · consistency 100
yes 100%
x-ai/grok-4.20 (20 runs) · consistency 100
yes 100%
x-ai/grok-4-fast (5 runs)
yes 100%
x-ai/grok-4.1-fast (10 runs) · consistency 100
yes 100%
x-ai/grok-4.20-beta (10 runs) · consistency 100
yes 100%
Xiaomi
xiaomi/mimo-v2.5 (20 runs) · consistency 44.4
yes 40%
hedge 55%
xiaomi/mimo-v2.5-pro (20 runs) · consistency 61.1
yes 80%
refusal 15%
xiaomi/mimo-v2-omni (15 runs) · consistency 77.8
yes 87%
xiaomi/mimo-v2-pro (15 runs) · consistency 61.1
yes 67%
no 13%
hedge 20%
Z.ai
z-ai/glm-5.2 (20 runs) · consistency 27.8
yes 25%
hedge 50%
refusal 25%
z-ai/glm-5.1 (20 runs) · consistency 44.4
yes 50%
refusal 50%
z-ai/glm-5-turbo (10 runs) · consistency 100
yes 100%
z-ai/glm-5 (10 runs) · consistency 44.4
other 60%
refusal 40%
z-ai/glm-4.7-flash (20 runs) · consistency 100
yes 100%
z-ai/glm-4.7 (14 runs) · consistency 61.1
yes 79%
refusal 21%
No models match.