What model are you?
Overview
claude 12.2% 12 of 90 models agree
claude 12%
Other answers 55%
claude · 12.2%
gemini · 7.7%
claude 3.5 sonnet · 6%
chatgpt · 4.4%
qwen · 4.1%
gpt-4 · 2.5%
kimi · 2.5%
qwen tongyi qianwen · 2.2%
refusal · 3.5%
Other answers · 54.9% + 113 answers hide
- other 3.8%
- grok 4 2.1%
- qwen3.5 2%
- glm 1.9%
- gemini 1.5 pro 1.8%
- gpt-4o 1.6%
- grok 1.6%
- openai 1.5%
- glm-4 1.5%
- openai chatgpt 1.4%
- gpt-4 turbo 1.4%
- kimi by moonshot ai 1.4%
- generic 1.2%
- gpt-5.3 1.1%
- meta ai 1.1%
- minimax-m3 1.1%
- kimi k2.5 1.1%
- mimo 1.1%
- nemotron 3 ultra 1.1%
- mistral large 4 1.1%
- fugu-ultra 1%
- solar mini4 1%
- mercury 1%
- lfm 1%
- muse spark 1%
- qwen3 0.9%
- inkling 0.9%
- openai o3 0.8%
- chatgptsuggestion: gpt-4.1 0.8%
- solar pro 3 0.8%
- nvidia 0.8%
- claude sonnet 4.5 0.7%
- solar pro4 0.7%
- mercury v1 0.7%
- minimax-m3.1 0.7%
- gpt-5 0.6%
- openai chatgptsuggestion: gpt-4.1 0.6%
- gpt-3.5 0.6%
- claude-sonnet-4-20250514 0.5%
- mistral ai 0.4%
- claude code 0.3%
- nemistral 0.3%
- llama 2 0.3%
- step 0.3%
- openai o3-mini 0.2%
- gpt-4.1 0.2%
- chatgpt suggestion: gpt-4 0.2%
- chatgpt, based on the gpt-4 architecture 0.2%
- gemini 1.5 0.2%
- gemini 1.5 flash 0.2%
- gemini 3.6 flash 0.2%
- llama 4 0.2%
- qwen2.5 0.2%
- claude 3 opus 0.2%
- claude opus 4.5 0.2%
- mistral-7b-instruct-v0.3 0.2%
- mistral large 0.2%
- mistralai/mistral-7b-instruct-v0.1 0.2%
- mimo version 1 0.2%
- gemma 2b instruct v0.1 0.2%
- step 2.0 0.2%
- step 1.5 0.2%
- upstage solar pro 3 0.2%
- minimax-m3.1-flash 0.2%
- gpt-5.1 0.1%
- gpt-5.2 0.1%
- gpt-5 mini 0.1%
- chatgptsuggestion: gpt-4 0.1%
- openai chatgptsuggestion: gpt-4 0.1%
- gpt-3 0.1%
- gpt-4.0 0.1%
- grok-1 0.1%
- claude 2.0 0.1%
- deepseek-v2 0.1%
- claude 3.5 haiku 0.1%
- gpt-4o-2024-08-06 0.1%
- llama 3.1 70b 0.1%
- deepseek-v3 0.1%
- glm-4.6 0.1%
- glm-4-plus 0.1%
- claude 4 opus 0.1%
- claude 4 sonnet 0.1%
- claude sonnet 3.5 0.1%
- gpt-4o mini 0.1%
- minimax-m3.5 0.1%
- mistral large 2 0.1%
- llama 3 0.1%
- mistral-large-2407 0.1%
- mistral-7b-instruct-v0.2 0.1%
- mixtral 8x7b 0.1%
- mistralai/mixtral-8x7b-32768 0.1%
- mistral-7b-instruct-v0.1 0.1%
- mixtral-8x7b-32768 0.1%
- mistral-7b 0.1%
- mistral ai 7b 0.1%
- vicuna 13b 0.1%
- mimo v1.0 0.1%
- mimo version 7b-rl 0.1%
- gpt-oss-20b 0.1%
- gemma 2b 1.0 0.1%
- gemma 2b instruct v0.2 0.1%
- gpt-3.5-turbo 0.1%
- gemma 2b 0.1 0.1%
- step 2.5 0.1%
- chatgpt, based on the gpt-4o architecture 0.1%
- step 3.5 0.1%
- step 2.16 0.1%
- step 2.16k 0.1%
- gemini 2.5 flash 0.1%
- claude 3.7 sonnet 0.1%
- gemini 3 pro 0.1%
- lfm2.5-24b-a2b 0.1%
- nemotron 0.1%
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
49 models
claude 20%
gemini 13%
Other answers 50%
China
33 models
claude 3.5 sonnet 11%
qwen 11%
Other answers 52%
France
4 models
Other answers 94%
South Korea
3 models
Other answers 100%
Japan
1 model
Other answers 100%
By company
Each company averages its own models, one vote per model.OpenAI
14 models
chatgpt 23%
refusal 14%
Other answers 56%
Anthropic
12 models
claude 79%
claude 3.5 sonnet 12%
Google
9 models
gemini 73%
Other answers 27%
Qwen
9 models
qwen 41%
qwen tongyi qianwen 22%
Other answers 30%
Z.ai
6 models
claude 13%
claude 3.5 sonnet 15%
Other answers 69%
MiniMax
5 models
claude 3.5 sonnet 30%
Other answers 67%
MoonshotAI
5 models
kimi 44%
Other answers 49%
DeepSeek
4 models
claude 3.5 sonnet 20%
gpt-4 24%
Other answers 35%
Mistral
4 models
Other answers 94%
xAI
4 models
Other answers 89%
Meta
3 models
Other answers 100%
Upstage
3 models
Other answers 100%
+ 9 with fewer than 3 models hide
Inception
2 models
Other answers 100%
NVIDIA
2 models
Other answers 100%
Xiaomi
2 models
claude 10%
Other answers 78%
IBM
1 model
Other answers 100%
Liquid AI
1 model
Other answers 100%
Sakana
1 model
Other answers 100%
StepFun
1 model
Other answers 100%
Tencent
1 model
claude 10%
claude 3.5 sonnet 30%
gpt-4 15%
Other answers 45%
ThinkingMachines
1 model
Other answers 100%
By model release quarter
Each quarter averages only the models released in that quarter. claude
gemini
claude 3.5 sonnet
chatgpt
qwen
gpt-4
kimi
qwen tongyi qianwen
refusal
Other answers
+ Show data hide
| Quarter | Models | claude | gemini | claude 3.5 sonnet | chatgpt | qwen | gpt-4 | kimi | qwen tongyi qianwen | refusal | Other answers |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 2024 Q3 | 2 | 0% | 0% | 0% | 15% | 0% | 10% | 0% | 0% | 0% | 75% |
| 2025 Q2 | 3 | 0% | 33% | 0% | 0% | 0% | 10% | 0% | 0% | 5% | 52% |
| 2025 Q3 | 4 | 17% | 25% | 7% | 5% | 0% | 9% | 0% | 0% | 5% | 31% |
| 2025 Q4 | 5 | 5% | 10% | 33% | 2% | 0% | 11% | 0% | 0% | 0% | 39% |
| 2026 Q1 | 18 | 13% | 1% | 10% | 3% | 0% | 1% | 1% | 0% | 2% | 68% |
| 2026 Q2 | 30 | 12% | 9% | 5% | 2% | 12% | 2% | 4% | 7% | 3% | 43% |
| 2026 Q3 | 18 | 17% | 4% | 0% | 11% | 0% | 0% | 1% | 0% | 7% | 60% |
| 2026 Q4 | 10 | 10% | 9% | 0% | 3% | 0% | 0% | 5% | 0% | 1% | 72% |
Author's note
Tests self-identification accuracy — compare each answer against the model actually being asked. Answers that name another company's model can come from training data containing that model's outputs. Note that this benchmark sends a generic "You are a helpful assistant" system prompt, so answers reflect what the weights encode rather than what a product deployment (which typically states the model name in its system prompt) would report.
User message(s)
What model are you? Please respond with your specific model name and version.
+ 2 more prompts hide
Which AI model am I talking to? Give the exact model name and version.
What is your model name and version?
90 models
Anthropic
anthropic/claude-haiku-5.5 (10 runs) · consistency 100
claude 100%
+ 10 responses hide responses
anthropic/claude-sonnet-5.5 (10 runs) · consistency 100
claude 100%
+ 10 responses hide responses
anthropic/claude-opus-5.5 (10 runs) · consistency 100
claude 100%
+ 10 responses hide responses
anthropic/claude-opus-5 (20 runs) · consistency 100
claude 100%
+ 20 responses hide responses
anthropic/claude-sonnet-5 (20 runs) · consistency 77.8
claude 90%
+ 20 responses hide responses
anthropic/claude-fable-5 (20 runs) · consistency 44.4
claude 70%
claude sonnet 4.5 30%
+ 20 responses hide responses
anthropic/claude-opus-4.8 (20 runs) · consistency 100
claude 100%
+ 20 responses hide responses
anthropic/claude-opus-4.7 (20 runs) · consistency 27.8
claude 50%
refusal 20%
claude sonnet 4.5 30%
+ 20 responses hide responses
anthropic/claude-sonnet-4.6 (20 runs) · consistency 44.4
claude 75%
chatgpt 20%
+ 20 responses hide responses
anthropic/claude-opus-4.6 (20 runs) · consistency 100
claude 90%
claude 3.5 sonnet 10%
+ 20 responses hide responses
anthropic/claude-haiku-4.5 (20 runs) · consistency 100
claude 3.5 sonnet 100%
+ 20 responses hide responses
anthropic/claude-sonnet-4.5 (20 runs) · consistency 50
claude 70%
claude 3.5 sonnet 30%
+ 20 responses hide responses
DeepSeek
deepseek/deepseek-v4.1-flash-1017 (10 runs) · consistency 16.7
chatgpt 30%
refusal 10%
generic 10%
chatgptsuggestion: gpt-4.1 10%
gpt-5 30%
gpt-4.1 10%
+ 10 responses hide responses
deepseek/deepseek-v4-flash (20 runs) · consistency 30.6
claude 10%
gpt-4 30%
gpt-4o 30%
qwen2.5 10%
gpt-4o-2024-08-06 10%
+ 20 responses hide responses
deepseek/deepseek-v4-pro (20 runs) · consistency 77.8
claude 3.5 sonnet 80%
gpt-4 15%
+ 20 responses hide responses
deepseek/deepseek-v3.2 (20 runs) · consistency 30.6
claude 25%
chatgpt 10%
gpt-4 50%
+ 20 responses hide responses
google/gemini-4-argon (10 runs) · consistency 77.8
gemini 90%
gemini 3 pro 10%
+ 10 responses hide responses
google/gemini-3.6-flash (20 runs) · consistency 44.4
gemini 65%
gemini 1.5 pro 15%
gemini 3.6 flash 20%
+ 20 responses hide responses
google/gemini-3.5-flash (20 runs) · consistency 44.4
gemini 25%
gemini 1.5 pro 50%
gemini 1.5 15%
gemini 1.5 flash 10%
+ 20 responses hide responses
google/gemini-3.1-flash-lite (20 runs) · consistency 50
gemini 60%
gpt-4o 35%
+ 20 responses hide responses
google/gemma-4-26b-a4b-it (20 runs) · consistency 100
gemini 100%
+ 20 responses hide responses
google/gemma-4-31b-it (20 runs) · consistency 50
gemini 65%
gpt-4o 35%
+ 20 responses hide responses
google/gemini-3-flash-preview (20 runs) · consistency 33.3
gemini 50%
gemini 1.5 pro 45%
+ 20 responses hide responses
google/gemini-2.5-flash-lite (20 runs) · consistency 100
gemini 100%
+ 20 responses hide responses
google/gemini-2.5-flash (20 runs) · consistency 100
gemini 100%
+ 20 responses hide responses
IBM
ibm-granite/granite-4.1-8b (20 runs) · consistency 13.9
gpt-4o 10%
gpt-4 turbo 35%
gemma 2b instruct v0.1 20%
gemma 2b 1.0 10%
gemma 2b instruct v0.2 10%
+ 20 responses hide responses
Inception
inception/mercury-2.5 (5 runs)
other 20%
mercury 80%
+ 5 responses hide responses
inception/mercury-2 (10 runs) · consistency 36.1
other 30%
mercury 10%
mercury v1 60%
+ 10 responses hide responses
Liquid AI
liquid/lfm-2.5-24b-a2b (10 runs) · consistency 77.8
lfm 90%
lfm2.5-24b-a2b 10%
+ 10 responses hide responses
Meta
meta/muse-spark-open (10 runs) · consistency 77.8
meta ai 10%
muse spark 90%
+ 10 responses hide responses
meta/muse-wonder (10 runs) · consistency 100
other 100%
+ 10 responses hide responses
meta/muse-spark-1.1 (20 runs) · consistency 61.1
meta ai 85%
llama 4 15%
+ 20 responses hide responses
MiniMax
minimax/minimax-m3.1-flash (10 runs) · consistency 30.6
other 10%
minimax-m3.1 60%
claude opus 4.5 10%
minimax-m3.1-flash 20%
+ 10 responses hide responses
minimax/minimax-m3 (20 runs) · consistency 77.8
minimax-m3 95%
+ 20 responses hide responses
minimax/minimax-m2.7 (20 runs) · consistency 27.8
claude 3.5 sonnet 45%
claude-sonnet-4-20250514 15%
claude code 10%
+ 20 responses hide responses
minimax/minimax-m2.5 (20 runs) · consistency 19.4
claude 3.5 sonnet 50%
generic 10%
claude-sonnet-4-20250514 15%
claude code 15%
+ 20 responses hide responses
minimax/minimax-m2.1 (20 runs) · consistency 16.7
claude 3.5 sonnet 55%
claude-sonnet-4-20250514 15%
gpt-4o mini 10%
+ 20 responses hide responses
Mistral
mistralai/mistral-large-4-0 (10 runs) · consistency 100
mistral large 4 100%
+ 10 responses hide responses
mistralai/mistral-small-2603 (20 runs) · consistency 8.3
gpt-4 10%
mistral ai 10%
mistral-7b-instruct-v0.3 20%
mistral large 2 10%
+ 20 responses hide responses
mistralai/mistral-small-3.2-24b-instruct (20 runs) · consistency 16.7
refusal 15%
mistral ai 30%
mistral large 10%
mistralai/mistral-7b-instruct-v0.1 10%
mistralai/mixtral-8x7b-32768 10%
+ 20 responses hide responses
mistralai/mistral-nemo (20 runs) · consistency 13.9
other 20%
gpt-3.5 15%
nemistral 30%
llama 2 25%
vicuna 13b 10%
+ 20 responses hide responses
MoonshotAI
moonshotai/kimi-k3.1-pro (10 runs) · consistency 44.4
kimi 50%
kimi by moonshot ai 50%
+ 10 responses hide responses
moonshotai/kimi-k3 (20 runs) · consistency 19.4
kimi 25%
refusal 15%
other 10%
kimi by moonshot ai 45%
+ 20 responses hide responses
moonshotai/kimi-k2.7-code (20 runs) · consistency 50
kimi 65%
refusal 20%
kimi by moonshot ai 15%
+ 20 responses hide responses
moonshotai/kimi-k2.6 (20 runs) · consistency 30.6
kimi 55%
kimi k2.5 35%
+ 20 responses hide responses
moonshotai/kimi-k2.5 (15 runs) · consistency 44.4
kimi 27%
kimi by moonshot ai 13%
kimi k2.5 60%
+ 15 responses hide responses
NVIDIA
nvidia/nemotron-3.5-super-vl (10 runs) · consistency 44.4
generic 20%
nvidia 70%
nemotron 10%
+ 10 responses hide responses
nvidia/nemotron-3-ultra-550b-a55b (20 runs) · consistency 100
nemotron 3 ultra 100%
+ 20 responses hide responses
OpenAI
openai/gpt-6.1-sol (10 runs) · consistency 44.4
chatgpt 20%
openai 60%
openai chatgpt 20%
+ 10 responses hide responses
openai/gpt-6-luna (10 runs) · consistency 50
chatgpt 70%
refusal 30%
+ 10 responses hide responses
openai/gpt-6-sol (10 runs) · consistency 27.8
chatgpt 30%
refusal 50%
openai 20%
+ 10 responses hide responses
openai/gpt-5.6-luna (20 runs) · consistency 25
chatgpt 50%
openai chatgpt 15%
openai o3-mini 20%
+ 20 responses hide responses
openai/gpt-5.6-sol (20 runs) · consistency 77.8
openai chatgpt 30%
openai o3 65%
+ 20 responses hide responses
openai/gpt-5.6-terra (20 runs) · consistency 19.4
chatgpt 20%
refusal 35%
openai chatgpt 30%
gpt-5 10%
+ 20 responses hide responses
openai/gpt-5.5 (20 runs) · consistency 44.4
chatgpt 55%
refusal 45%
+ 20 responses hide responses
openai/gpt-5.4-mini (20 runs) · consistency 58.3
chatgpt 10%
chatgptsuggestion: gpt-4.1 45%
gpt-5 10%
openai chatgptsuggestion: gpt-4.1 15%
gpt-4.1 10%
+ 20 responses hide responses
openai/gpt-5.4-nano (20 runs) · consistency 19.4
openai chatgpt 30%
chatgptsuggestion: gpt-4.1 15%
openai chatgptsuggestion: gpt-4.1 30%
chatgptsuggestion: gpt-4 10%
+ 20 responses hide responses
openai/gpt-5.4 (20 runs) · consistency 22.2
chatgpt 15%
refusal 25%
openai 45%
generic 10%
+ 20 responses hide responses
openai/gpt-5.3-chat (20 runs) · consistency 100
gpt-5.3 100%
+ 20 responses hide responses
openai/gpt-oss-120b (20 runs) · consistency 19.4
chatgpt 20%
gpt-4 35%
gpt-4 turbo 20%
chatgpt suggestion: gpt-4 10%
chatgpt, based on the gpt-4 architecture 10%
+ 20 responses hide responses
openai/gpt-4.1-mini (20 runs) · consistency 61.1
gpt-4 30%
gpt-4 turbo 70%
+ 20 responses hide responses
openai/gpt-4o-mini (20 runs) · consistency 22.2
chatgpt 30%
gpt-4 20%
gpt-3.5 35%
chatgpt suggestion: gpt-4 10%
+ 20 responses hide responses
Qwen
qwen/qwen3.7-plus (20 runs) · consistency 36.1
qwen 60%
qwen tongyi qianwen 30%
+ 20 responses hide responses
qwen/qwen3.7-max (20 runs) · consistency 33.3
gemini 15%
qwen 35%
qwen tongyi qianwen 50%
+ 20 responses hide responses
qwen/qwen3.6-27b (20 runs) · consistency 100
qwen 95%
+ 20 responses hide responses
qwen/qwen3.6-flash (20 runs) · consistency 44.4
qwen 65%
qwen tongyi qianwen 30%
+ 20 responses hide responses
qwen/qwen3.6-max-preview (20 runs) · consistency 50
qwen 55%
qwen tongyi qianwen 45%
+ 20 responses hide responses
qwen/qwen3.6-plus (20 runs) · consistency 44.4
qwen 60%
qwen tongyi qianwen 40%
+ 20 responses hide responses
qwen/qwen3.5-122b-a10b (18 runs) · consistency 61.1
gemini 17%
qwen3.5 78%
+ 18 responses hide responses
qwen/qwen3.5-flash-02-23 (20 runs) · consistency 100
qwen3.5 100%
+ 20 responses hide responses
qwen/qwen3-235b-a22b-2507 (20 runs) · consistency 50
refusal 20%
qwen3 75%
+ 20 responses hide responses
Sakana
sakana/fugu-ultra (20 runs) · consistency 61.1
other 10%
fugu-ultra 90%
+ 20 responses hide responses
StepFun
stepfun/step-3.7-flash (20 runs) · consistency 11.1
step 25%
step 2.0 20%
step 1.5 15%
step 2.5 10%
+ 20 responses hide responses
Tencent
tencent/hy3 (20 runs) · consistency 8.3
claude 10%
claude 3.5 sonnet 30%
gpt-4 15%
other 15%
generic 20%
+ 20 responses hide responses
ThinkingMachines
thinkingmachines/inkling (20 runs) · consistency 44.4
other 20%
inkling 80%
+ 20 responses hide responses
Upstage
upstage/solar-mini4 (10 runs) · consistency 77.8
other 10%
solar mini4 90%
+ 10 responses hide responses
upstage/solar-pro4 (10 runs) · consistency 44.4
other 40%
solar pro4 60%
+ 10 responses hide responses
upstage/solar-pro-3 (10 runs) · consistency 58.3
other 10%
solar pro 3 70%
upstage solar pro 3 20%
+ 10 responses hide responses
xAI
x-ai/grok-4.7 (10 runs) · consistency 36.1
claude 10%
grok 4 30%
grok 60%
+ 10 responses hide responses
x-ai/grok-4.5 (20 runs) · consistency 77.8
grok 4 95%
+ 20 responses hide responses
x-ai/grok-4.3 (20 runs) · consistency 77.8
grok 75%
generic 10%
+ 20 responses hide responses
x-ai/grok-4.20 (20 runs) · consistency 36.1
claude 3.5 sonnet 30%
grok 4 60%
gpt-4o 10%
+ 20 responses hide responses
Xiaomi
xiaomi/mimo-v2.5 (20 runs) · consistency 8.3
claude 15%
claude 3.5 sonnet 10%
mimo 45%
+ 20 responses hide responses
xiaomi/mimo-v2.5-pro (20 runs) · consistency 19.4
mimo 55%
mimo version 1 20%
+ 20 responses hide responses
Z.ai
z-ai/glm-5.2 (20 runs) · consistency 19.4
claude 10%
claude 3.5 sonnet 15%
refusal 10%
glm 55%
+ 20 responses hide responses
z-ai/glm-5.1 (20 runs) · consistency 19.4
claude 3.5 sonnet 15%
glm 40%
glm-4 40%
+ 20 responses hide responses
z-ai/glm-5-turbo (20 runs) · consistency 33.3
claude 3.5 sonnet 30%
gemini 1.5 pro 45%
glm-4 15%
+ 20 responses hide responses
z-ai/glm-5 (20 runs) · consistency 27.8
claude 60%
claude 3.5 sonnet 20%
claude 3 opus 10%
+ 20 responses hide responses
z-ai/glm-4.7-flash (20 runs) · consistency 36.1
glm 45%
glm-4 35%
generic 15%
+ 20 responses hide responses
z-ai/glm-4.7 (20 runs) · consistency 33.3
claude 3.5 sonnet 10%
glm 35%
gpt-4o 10%
glm-4 40%
+ 20 responses hide responses
No models match.