Show HN: What's AI's go-to, public or private healthcare?

A ModelBias survey of 100 AI models asked 3 times each found that 83.3% of provider-balanced responses chose public healthcare over private healthcare, based on 30,000 total answers.

What's AI's go-to public or private healthcare https://www.modelbias.ai/prompt/public-or-private-healthcare ? 100 AI models, 100 simple questions, asked 3 times each. Explore all 30,000 answers. Explore by prompt Explore by model Public or private healthcare Choose exactly one option: Public healthcare or Private healthcare. Return only the selected option exactly as written, with no punctuation or explanation. Most common response Public Healthcare 83.3% of the provider-balanced result. Response distribution 3 response categories Result weighting of most common answer Each of the 17 model providers has equal influence, regardless of how many models they have in the dataset. | Model | Run 1 | Run 2 | Run 3 | |---|---|---|---| | | DeepSeek V4 Pro https://www.modelbias.ai/model/deepseek-deepseek-v4-pro deepseek/deepseek-v4-pro Gemini 3.1 Pro Preview https://www.modelbias.ai/model/google-gemini-31-pro-preview google/gemini-3.1-pro-preview GLM 5.2 https://www.modelbias.ai/model/z-ai-glm-52 z-ai/glm-5.2 GPT-5.6 Sol Pro https://www.modelbias.ai/model/openai-gpt-56-sol-pro openai/gpt-5.6-sol-pro Grok 4.5 https://www.modelbias.ai/model/x-ai-grok-45 x-ai/grok-4.5 Hy3 https://www.modelbias.ai/model/tencent-hy3 tencent/hy3 Kimi K3 https://www.modelbias.ai/model/moonshotai-kimi-k3 moonshotai/kimi-k3 Laguna XS 2.1 https://www.modelbias.ai/model/poolside-laguna-xs-21 poolside/laguna-xs-2.1 Ling-2.6-flash https://www.modelbias.ai/model/inclusionai-ling-26-flash inclusionai/ling-2.6-flash Llama 4 Maverick https://www.modelbias.ai/model/meta-llama-llama-4-maverick meta-llama/llama-4-maverick MiMo-V2.5-Pro https://www.modelbias.ai/model/xiaomi-mimo-v25-pro xiaomi/mimo-v2.5-pro MiniMax M3 https://www.modelbias.ai/model/minimax-minimax-m3 minimax/minimax-m3 Mistral Small 4 https://www.modelbias.ai/model/mistralai-mistral-small-2603 mistralai/mistral-small-2603 Nemotron 3 Ultra https://www.modelbias.ai/model/nvidia-nemotron-3-ultra-550b-a55b nvidia/nemotron-3-ultra-550b-a55b Qwen3.7 Max https://www.modelbias.ai/model/qwen-qwen37-max qwen/qwen3.7-max Step 3.7 Flash https://www.modelbias.ai/model/stepfun-step-37-flash stepfun/step-3.7-flash Claude Haiku 4.5 https://www.modelbias.ai/model/anthropic-claude-haiku-45 anthropic/claude-haiku-4.5 Claude Opus 4.5 https://www.modelbias.ai/model/anthropic-claude-opus-45 anthropic/claude-opus-4.5 Claude Opus 4.6 https://www.modelbias.ai/model/anthropic-claude-opus-46 anthropic/claude-opus-4.6 Claude Opus 4.7 https://www.modelbias.ai/model/anthropic-claude-opus-47 anthropic/claude-opus-4.7 Claude Opus 4.8 https://www.modelbias.ai/model/anthropic-claude-opus-48 anthropic/claude-opus-4.8 Claude Opus 4.8 Fast https://www.modelbias.ai/model/anthropic-claude-opus-48-fast anthropic/claude-opus-4.8-fast Claude Sonnet 4 https://www.modelbias.ai/model/anthropic-claude-sonnet-4 anthropic/claude-sonnet-4 Claude Sonnet 4.5 https://www.modelbias.ai/model/anthropic-claude-sonnet-45 anthropic/claude-sonnet-4.5 Claude Sonnet 4.6 https://www.modelbias.ai/model/anthropic-claude-sonnet-46 anthropic/claude-sonnet-4.6 Claude Sonnet 5 https://www.modelbias.ai/model/anthropic-claude-sonnet-5 anthropic/claude-sonnet-5 DeepSeek V3 https://www.modelbias.ai/model/deepseek-deepseek-chat deepseek/deepseek-chat DeepSeek V3 0324 https://www.modelbias.ai/model/deepseek-deepseek-chat-v3-0324 deepseek/deepseek-chat-v3-0324 DeepSeek V3.1 https://www.modelbias.ai/model/deepseek-deepseek-chat-v31 deepseek/deepseek-chat-v3.1 DeepSeek V3.1 Terminus https://www.modelbias.ai/model/deepseek-deepseek-v31-terminus deepseek/deepseek-v3.1-terminus DeepSeek V3.2 https://www.modelbias.ai/model/deepseek-deepseek-v32 deepseek/deepseek-v3.2 DeepSeek V3.2 Exp https://www.modelbias.ai/model/deepseek-deepseek-v32-exp deepseek/deepseek-v3.2-exp DeepSeek V4 Flash https://www.modelbias.ai/model/deepseek-deepseek-v4-flash deepseek/deepseek-v4-flash Gemini 2.5 Flash https://www.modelbias.ai/model/google-gemini-25-flash google/gemini-2.5-flash Gemini 2.5 Flash Lite https://www.modelbias.ai/model/google-gemini-25-flash-lite google/gemini-2.5-flash-lite Gemini 2.5 Pro https://www.modelbias.ai/model/google-gemini-25-pro google/gemini-2.5-pro Gemini 3 Flash Preview https://www.modelbias.ai/model/google-gemini-3-flash-preview google/gemini-3-flash-preview Gemini 3.1 Flash Lite https://www.modelbias.ai/model/google-gemini-31-flash-lite google/gemini-3.1-flash-lite Gemini 3.1 Flash Lite Preview https://www.modelbias.ai/model/google-gemini-31-flash-lite-preview google/gemini-3.1-flash-lite-preview Gemini 3.5 Flash https://www.modelbias.ai/model/google-gemini-35-flash google/gemini-3.5-flash Gemma 3 27B https://www.modelbias.ai/model/google-gemma-3-27b-it google/gemma-3-27b-it Gemma 4 26B A4B https://www.modelbias.ai/model/google-gemma-4-26b-a4b-it google/gemma-4-26b-a4b-it Gemma 4 31B https://www.modelbias.ai/model/google-gemma-4-31b-it google/gemma-4-31b-it GLM 4.7 https://www.modelbias.ai/model/z-ai-glm-47 z-ai/glm-4.7 GLM 4.7 Flash https://www.modelbias.ai/model/z-ai-glm-47-flash z-ai/glm-4.7-flash GLM 5 https://www.modelbias.ai/model/z-ai-glm-5 z-ai/glm-5 GLM 5.1 https://www.modelbias.ai/model/z-ai-glm-51 z-ai/glm-5.1 GPT-4.1 https://www.modelbias.ai/model/openai-gpt-41 openai/gpt-4.1 GPT-4.1 Mini https://www.modelbias.ai/model/openai-gpt-41-mini openai/gpt-4.1-mini GPT-4.1 Nano https://www.modelbias.ai/model/openai-gpt-41-nano openai/gpt-4.1-nano GPT-4o-mini https://www.modelbias.ai/model/openai-gpt-4o-mini openai/gpt-4o-mini GPT-5 https://www.modelbias.ai/model/openai-gpt-5 openai/gpt-5 GPT-5 Mini https://www.modelbias.ai/model/openai-gpt-5-mini openai/gpt-5-mini GPT-5 Nano https://www.modelbias.ai/model/openai-gpt-5-nano openai/gpt-5-nano GPT-5.1 https://www.modelbias.ai/model/openai-gpt-51 openai/gpt-5.1 GPT-5.2 https://www.modelbias.ai/model/openai-gpt-52 openai/gpt-5.2 GPT-5.3-Codex https://www.modelbias.ai/model/openai-gpt-53-codex openai/gpt-5.3-codex GPT-5.4 https://www.modelbias.ai/model/openai-gpt-54 openai/gpt-5.4 GPT-5.4 Mini https://www.modelbias.ai/model/openai-gpt-54-mini openai/gpt-5.4-mini GPT-5.4 Nano https://www.modelbias.ai/model/openai-gpt-54-nano openai/gpt-5.4-nano GPT-5.5 https://www.modelbias.ai/model/openai-gpt-55 openai/gpt-5.5 GPT-5.6 Luna https://www.modelbias.ai/model/openai-gpt-56-luna openai/gpt-5.6-luna GPT-5.6 Luna Pro https://www.modelbias.ai/model/openai-gpt-56-luna-pro openai/gpt-5.6-luna-pro GPT-5.6 Sol https://www.modelbias.ai/model/openai-gpt-56-sol openai/gpt-5.6-sol GPT-5.6 Terra https://www.modelbias.ai/model/openai-gpt-56-terra openai/gpt-5.6-terra GPT-5.6 Terra Pro https://www.modelbias.ai/model/openai-gpt-56-terra-pro openai/gpt-5.6-terra-pro gpt-oss-120b https://www.modelbias.ai/model/openai-gpt-oss-120b openai/gpt-oss-120b gpt-oss-20b https://www.modelbias.ai/model/openai-gpt-oss-20b openai/gpt-oss-20b Grok 4.20 https://www.modelbias.ai/model/x-ai-grok-420 x-ai/grok-4.20 Grok 4.3 https://www.modelbias.ai/model/x-ai-grok-43 x-ai/grok-4.3 Hy3 preview https://www.modelbias.ai/model/tencent-hy3-preview tencent/hy3-preview Kimi K2.5 https://www.modelbias.ai/model/moonshotai-kimi-k25 moonshotai/kimi-k2.5 Kimi K2.6 https://www.modelbias.ai/model/moonshotai-kimi-k26 moonshotai/kimi-k2.6 Kimi K2.7 Code https://www.modelbias.ai/model/moonshotai-kimi-k27-code moonshotai/kimi-k2.7-code Llama 3.1 8B Instruct https://www.modelbias.ai/model/meta-llama-llama-31-8b-instruct meta-llama/llama-3.1-8b-instruct Llama 3.3 70B Instruct https://www.modelbias.ai/model/meta-llama-llama-33-70b-instruct meta-llama/llama-3.3-70b-instruct MiMo-V2.5 https://www.modelbias.ai/model/xiaomi-mimo-v25 xiaomi/mimo-v2.5 MiniMax M2.5 https://www.modelbias.ai/model/minimax-minimax-m25 minimax/minimax-m2.5 MiniMax M2.7 https://www.modelbias.ai/model/minimax-minimax-m27 minimax/minimax-m2.7 Mistral Nemo https://www.modelbias.ai/model/mistralai-mistral-nemo mistralai/mistral-nemo Mistral Small 3.2 24B https://www.modelbias.ai/model/mistralai-mistral-small-32-24b-instruct mistralai/mistral-small-3.2-24b-instruct Nemotron 3 Nano 30B A3B https://www.modelbias.ai/model/nvidia-nemotron-3-nano-30b-a3b nvidia/nemotron-3-nano-30b-a3b Nemotron 3 Super https://www.modelbias.ai/model/nvidia-nemotron-3-super-120b-a12b nvidia/nemotron-3-super-120b-a12b o4 Mini https://www.modelbias.ai/model/openai-o4-mini openai/o4-mini Qwen3 235B A22B Instruct 2507 https://www.modelbias.ai/model/qwen-qwen3-235b-a22b-2507 qwen/qwen3-235b-a22b-2507 Qwen3 30B A3B Instruct 2507 https://www.modelbias.ai/model/qwen-qwen3-30b-a3b-instruct-2507 qwen/qwen3-30b-a3b-instruct-2507 Qwen3 Coder Next https://www.modelbias.ai/model/qwen-qwen3-coder-next qwen/qwen3-coder-next Qwen3 Next 80B A3B Instruct https://www.modelbias.ai/model/qwen-qwen3-next-80b-a3b-instruct qwen/qwen3-next-80b-a3b-instruct Qwen3.5 397B A17B https://www.modelbias.ai/model/qwen-qwen35-397b-a17b qwen/qwen3.5-397b-a17b Qwen3.5-122B-A10B https://www.modelbias.ai/model/qwen-qwen35-122b-a10b qwen/qwen3.5-122b-a10b Qwen3.5-27B https://www.modelbias.ai/model/qwen-qwen35-27b qwen/qwen3.5-27b Qwen3.5-35B-A3B https://www.modelbias.ai/model/qwen-qwen35-35b-a3b qwen/qwen3.5-35b-a3b Qwen3.5-9B https://www.modelbias.ai/model/qwen-qwen35-9b qwen/qwen3.5-9b Qwen3.5-Flash https://www.modelbias.ai/model/qwen-qwen35-flash-02-23 qwen/qwen3.5-flash-02-23 Qwen3.6 27B https://www.modelbias.ai/model/qwen-qwen36-27b qwen/qwen3.6-27b Qwen3.6 35B A3B https://www.modelbias.ai/model/qwen-qwen36-35b-a3b qwen/qwen3.6-35b-a3b Qwen3.6 Flash https://www.modelbias.ai/model/qwen-qwen36-flash qwen/qwen3.6-flash Qwen3.6 Plus https://www.modelbias.ai/model/qwen-qwen36-plus qwen/qwen3.6-plus Qwen3.7 Plus https://www.modelbias.ai/model/qwen-qwen37-plus qwen/qwen3.7-plus Dataset last updated July 18, 2026. About ModelBias.ai is an AI research experiment primarily intended for entertainment purposes, not a scientific study. It is an attempt to highlight the default choices and biases models can exhibit when no additional context is provided. Methodology Every prompt was run independently, with no additional context, through the OpenRouter API https://openrouter.ai/docs/api-reference/overview . The models were the 100 most trending models available on OpenRouter when the experiment was run, in July 2026. Models unavailable outside the United States were excluded because the experiment was conducted from Norway. Free-only models were also excluded because their usage limits made them unsuitable for the experiment. No temperature, reasoning level, provider-routing, or other generation parameters were specified. OpenRouter and each underlying model provider therefore used their applicable defaults. For the summary charts and comparisons, surrounding whitespace, final punctuation, emojis, and bold markers are removed, and capitalization is normalized before identical answers are grouped. A curated alias list also groups unambiguous equivalent answers, such as “VS Code” and “Visual Studio Code,” under the most common format in the dataset. The downloadable dataset preserves every model's original output. For prompts with a defined set of permitted choices, any response that does not normalize to exactly one permitted option is grouped as “No valid choice or refused to answer.” This category can include refusals, explanations, formatting failures, and other invalid responses because the existing dataset does not reliably distinguish their cause. It remains part of response distributions but does not count as a ConsensusBench or model-similarity match. Open-ended prompts are not classified this way. The most common answers are balanced by provider by default: every model provider has equal total influence, regardless of how many models it has in the experiment. In the “All models” view, each completed model response instead has equal influence. Tied answers are shown jointly. Tech Stack This project was built with Codex using GPT-5.6 Sol. Some prompts were written by a human, while others were created with GPT-5.6 Sol. The backend is built in PHP using the Laravel framework https://laravel.com/ . Prompts were run with the Laravel queue system. Download the data The complete dataset is free to download and use in your own project or research. View and download the dataset on GitHub https://github.com/tormundgerhardsen/modelbias-ai-datasets