LLM Behavioral Fingerprint Database
Ask a model for a random number and it betrays itself: every LLM has stable favorite answers. This database shows the one-word answer distributions of 167 models across up to 15 probe cells (English and Chinese) at temperature 1.0, from the paper's open dataset. Click any model to see its full fingerprint.
Is your API endpoint really the model it claims?
Sample your endpoint in your browser and compare its distributions against these reference fingerprints. Your API key never leaves the page.
Open the fingerprint checker167 models
AI21(1)
Amazon(5)
Nova 2 Lite
amazon/nova-2-lite-v1 · 15 cells
Nova Lite 1.0
amazon/nova-lite-v1 · 15 cells
Nova Micro 1.0
amazon/nova-micro-v1 · 15 cells
Nova Premier 1.0
amazon/nova-premier-v1 · 15 cells
Nova Pro 1.0
amazon/nova-pro-v1 · 14 cells
Anthropic(12)
Claude 3 Haiku
anthropic/claude-3-haiku · 15 cells
Claude Haiku 4.5
anthropic/claude-haiku-4.5 · 14 cells
Claude Opus 4
anthropic/claude-opus-4 · 15 cells
Claude Opus 4.1
anthropic/claude-opus-4.1 · 15 cells
Claude Opus 4.5
anthropic/claude-opus-4.5 · 15 cells
Claude Opus 4.6
anthropic/claude-opus-4.6 · 15 cells
Claude Opus 4.7
anthropic/claude-opus-4.7 · 15 cells
Claude Opus 4.8
anthropic/claude-opus-4.8 · 15 cells
Claude Sonnet 4
anthropic/claude-sonnet-4 · 15 cells
Claude Sonnet 4.5
anthropic/claude-sonnet-4.5 · 15 cells
Claude Sonnet 4.6
anthropic/claude-sonnet-4.6 · 15 cells
Claude Sonnet 5
anthropic/claude-sonnet-5 · 15 cells
Baidu(1)
ByteDance Seed(4)
Seed 1.6
bytedance-seed/seed-1.6 · 15 cells
Seed 1.6 Flash
bytedance-seed/seed-1.6-flash · 15 cells
Seed-2.0-Lite
bytedance-seed/seed-2.0-lite · 15 cells
Seed-2.0-Mini
bytedance-seed/seed-2.0-mini · 15 cells
Cohere(4)
Command A
cohere/command-a · 15 cells
Command R (08-2024)
cohere/command-r-08-2024 · 15 cells
Command R+ (08-2024)
cohere/command-r-plus-08-2024 · 15 cells
Command R7B (12-2024)
cohere/command-r7b-12-2024 · 14 cells
Deep Cogito(1)
DeepSeek(8)
DeepSeek V3
deepseek/deepseek-chat · 15 cells
DeepSeek V3 0324
deepseek/deepseek-chat-v3-0324 · 15 cells
DeepSeek V3.1
deepseek/deepseek-chat-v3.1 · 15 cells
DeepSeek V3.1 Terminus
deepseek/deepseek-v3.1-terminus · 15 cells
DeepSeek V3.2
deepseek/deepseek-v3.2 · 15 cells
DeepSeek V3.2 Exp
deepseek/deepseek-v3.2-exp · 15 cells
DeepSeek V4 Flash
deepseek/deepseek-v4-flash · 15 cells
DeepSeek V4 Pro
deepseek/deepseek-v4-pro · 15 cells
Google(10)
Gemini 2.5 Flash
google/gemini-2.5-flash · 15 cells
Gemini 2.5 Flash Lite
google/gemini-2.5-flash-lite · 15 cells
Gemini 3.1 Flash Lite
google/gemini-3.1-flash-lite · 15 cells
Gemma 2 27B
google/gemma-2-27b-it · 15 cells
Gemma 3 12B
google/gemma-3-12b-it · 15 cells
Gemma 3 27B
google/gemma-3-27b-it · 15 cells
Gemma 3 4B
google/gemma-3-4b-it · 15 cells
Gemma 3n 4B
google/gemma-3n-e4b-it · 15 cells
Gemma 4 26B A4B
google/gemma-4-26b-a4b-it · 15 cells
Gemma 4 31B
google/gemma-4-31b-it · 15 cells
IBM(2)
Inception(1)
inclusionAI(2)
LiquidAI(1)
Meta(9)
Llama 3 8B Instruct
meta-llama/llama-3-8b-instruct · 15 cells
Llama 3.1 70B Instruct
meta-llama/llama-3.1-70b-instruct · 15 cells
Llama 3.1 8B Instruct
meta-llama/llama-3.1-8b-instruct · 15 cells
Llama 3.2 11B Vision Instruct
meta-llama/llama-3.2-11b-vision-instruct · 15 cells
Llama 3.2 1B Instruct
meta-llama/llama-3.2-1b-instruct · 15 cells
Llama 3.2 3B Instruct
meta-llama/llama-3.2-3b-instruct · 14 cells
Llama 3.3 70B Instruct
meta-llama/llama-3.3-70b-instruct · 15 cells
Llama 4 Maverick
meta-llama/llama-4-maverick · 15 cells
Llama 4 Scout
meta-llama/llama-4-scout · 15 cells
Microsoft(1)
microsoft(1)
MiniMax(2)
Mistral(15)
Ministral 3 14B 2512
mistralai/ministral-14b-2512 · 15 cells
Ministral 3 3B 2512
mistralai/ministral-3b-2512 · 14 cells
Ministral 3 8B 2512
mistralai/ministral-8b-2512 · 15 cells
Mistral Large 3 2512
mistralai/mistral-large-2512 · 15 cells
Mistral Medium 3
mistralai/mistral-medium-3 · 15 cells
Mistral Medium 3.1
mistralai/mistral-medium-3.1 · 15 cells
Mistral Medium 3.5
mistralai/mistral-medium-3-5 · 15 cells
Mistral Nemo
mistralai/mistral-nemo · 15 cells
Mistral Small 3
mistralai/mistral-small-24b-instruct-2501 · 15 cells
Mistral Small 3.1 24B
mistralai/mistral-small-3.1-24b-instruct · 9 cells
Mistral Small 3.2 24B
mistralai/mistral-small-3.2-24b-instruct · 15 cells
Mistral Small 4
mistralai/mistral-small-2603 · 15 cells
Mixtral 8x22B Instruct
mistralai/mixtral-8x22b-instruct · 15 cells
Saba
mistralai/mistral-saba · 15 cells
Voxtral Small 24B 2507
mistralai/voxtral-small-24b-2507 · 15 cells
mistralai(2)
MoonshotAI(4)
Kimi K2 0711
moonshotai/kimi-k2 · 15 cells
Kimi K2 0905
moonshotai/kimi-k2-0905 · 15 cells
Kimi K2.5
moonshotai/kimi-k2.5 · 15 cells
Kimi K2.6
moonshotai/kimi-k2.6 · 15 cells
Nous(4)
Hermes 3 405B Instruct
nousresearch/hermes-3-llama-3.1-405b · 15 cells
Hermes 3 70B Instruct
nousresearch/hermes-3-llama-3.1-70b · 15 cells
Hermes 4 405B
nousresearch/hermes-4-405b · 15 cells
Hermes 4 70B
nousresearch/hermes-4-70b · 15 cells
NVIDIA(4)
Llama 3.3 Nemotron Super 49B V1.5
nvidia/llama-3.3-nemotron-super-49b-v1.5 · 15 cells
Nemotron 3 Nano 30B A3B
nvidia/nemotron-3-nano-30b-a3b · 15 cells
Nemotron 3 Super
nvidia/nemotron-3-super-120b-a12b · 15 cells
Nemotron 3 Ultra
nvidia/nemotron-3-ultra-550b-a55b · 15 cells
OpenAI(21)
GPT-3.5 Turbo
openai/gpt-3.5-turbo · 15 cells
GPT-3.5 Turbo (older v0613)
openai/gpt-3.5-turbo-0613 · 15 cells
GPT-3.5 Turbo Instruct
openai/gpt-3.5-turbo-instruct · 14 cells
GPT-4
openai/gpt-4 · 15 cells
GPT-4 Turbo
openai/gpt-4-turbo · 15 cells
GPT-4.1
openai/gpt-4.1 · 14 cells
GPT-4.1 Mini
openai/gpt-4.1-mini · 15 cells
GPT-4.1 Nano
openai/gpt-4.1-nano · 15 cells
GPT-4o
openai/gpt-4o · 15 cells
GPT-4o (2024-05-13)
openai/gpt-4o-2024-05-13 · 15 cells
GPT-4o (2024-08-06)
openai/gpt-4o-2024-08-06 · 15 cells
GPT-4o (2024-11-20)
openai/gpt-4o-2024-11-20 · 15 cells
GPT-4o-mini
openai/gpt-4o-mini · 15 cells
GPT-4o-mini (2024-07-18)
openai/gpt-4o-mini-2024-07-18 · 15 cells
GPT-5 Chat
openai/gpt-5-chat · 15 cells
GPT-5.1
openai/gpt-5.1 · 15 cells
GPT-5.2
openai/gpt-5.2 · 15 cells
GPT-5.4
openai/gpt-5.4 · 15 cells
GPT-5.4 Mini
openai/gpt-5.4-mini · 15 cells
GPT-5.4 Nano
openai/gpt-5.4-nano · 15 cells
GPT-5.5
openai/gpt-5.5 · 15 cells
Perceptron(1)
Qwen(29)
Qwen Plus 0728
qwen/qwen-plus-2025-07-28 · 15 cells
Qwen-Plus
qwen/qwen-plus · 15 cells
Qwen2.5 7B Instruct
qwen/qwen-2.5-7b-instruct · 15 cells
Qwen2.5 VL 72B Instruct
qwen/qwen2.5-vl-72b-instruct · 15 cells
Qwen3 235B A22B
qwen/qwen3-235b-a22b · 15 cells
Qwen3 235B A22B Instruct 2507
qwen/qwen3-235b-a22b-2507 · 15 cells
Qwen3 30B A3B Instruct 2507
qwen/qwen3-30b-a3b-instruct-2507 · 15 cells
Qwen3 8B
qwen/qwen3-8b · 15 cells
Qwen3 Max
qwen/qwen3-max · 15 cells
Qwen3 Max Thinking
qwen/qwen3-max-thinking · 15 cells
Qwen3 Next 80B A3B Instruct
qwen/qwen3-next-80b-a3b-instruct · 15 cells
Qwen3 VL 235B A22B Instruct
qwen/qwen3-vl-235b-a22b-instruct · 15 cells
Qwen3 VL 30B A3B Instruct
qwen/qwen3-vl-30b-a3b-instruct · 15 cells
Qwen3 VL 32B Instruct
qwen/qwen3-vl-32b-instruct · 15 cells
Qwen3 VL 8B Instruct
qwen/qwen3-vl-8b-instruct · 15 cells
Qwen3.5 397B A17B
qwen/qwen3.5-397b-a17b · 15 cells
Qwen3.5 Plus 2026-02-15
qwen/qwen3.5-plus-02-15 · 15 cells
Qwen3.5 Plus 2026-04-20
qwen/qwen3.5-plus-20260420 · 15 cells
Qwen3.5-122B-A10B
qwen/qwen3.5-122b-a10b · 15 cells
Qwen3.5-27B
qwen/qwen3.5-27b · 15 cells
Qwen3.5-35B-A3B
qwen/qwen3.5-35b-a3b · 15 cells
Qwen3.5-9B
qwen/qwen3.5-9b · 15 cells
Qwen3.5-Flash
qwen/qwen3.5-flash-02-23 · 15 cells
Qwen3.6 27B
qwen/qwen3.6-27b · 15 cells
Qwen3.6 35B A3B
qwen/qwen3.6-35b-a3b · 15 cells
Qwen3.6 Flash
qwen/qwen3.6-flash · 15 cells
Qwen3.6 Plus
qwen/qwen3.6-plus · 15 cells
Qwen3.7 Max
qwen/qwen3.7-max · 15 cells
Qwen3.7 Plus
qwen/qwen3.7-plus · 15 cells
qwen(1)
rekaai(1)
Tencent(2)
Upstage(1)
Writer(1)
xAI(2)
Xiaomi(2)
Z.ai(12)
GLM 4.5
z-ai/glm-4.5 · 15 cells
GLM 4.5 Air
z-ai/glm-4.5-air · 15 cells
GLM 4.5V
z-ai/glm-4.5v · 15 cells
GLM 4.6
z-ai/glm-4.6 · 15 cells
GLM 4.6V
z-ai/glm-4.6v · 15 cells
GLM 4.7
z-ai/glm-4.7 · 15 cells
GLM 4.7 Flash
z-ai/glm-4.7-flash · 15 cells
GLM 5
z-ai/glm-5 · 15 cells
GLM 5 Turbo
z-ai/glm-5-turbo · 15 cells
GLM 5.1
z-ai/glm-5.1 · 15 cells
GLM 5.2
z-ai/glm-5.2 · 15 cells
GLM 5V Turbo
z-ai/glm-5v-turbo · 15 cells
Data source & license
Distributions: Tomáš Bruckner, "One Token Is Enough" (arXiv:2607.10252), official dataset Zenodo DOI 10.5281/zenodo.21278557, licensed CC-BY-4.0. Collected via OpenRouter at temperature 1.0 under the paper's fixed minimal one-word system prompt; answer keys are re-normalized with the checker's pipeline (color aliases, number words, coin h/t) so live probes are directly comparable.
Paper on arXivDataset on ZenodoCC-BY-4.0 license
Further reading
Deep dive: How to verify an LLM API is real — fingerprinting models with random numbers中文长文：AI 为什么说不出随机数？一枚「模型指纹」，照出 API 中转站的真伪