What's AI's go-to delayed project solution⁠?

100 AI models, 100 simple questions, asked 3 times each. Explore all 30,000 answers.

Explore by prompt

Explore by model

Delayed project solution

A project is behind schedule. Choose exactly one option: Reduce scope, Delay launch, or Add workers. Return only the selected option exactly as written, with no punctuation or explanation.

Most common response

Reduce Scope

57.8% of the provider-balanced result.

Response distribution

3 response categories

Reduce Scope 57.8%

Add Workers 26.5%

Delay Launch 15.7%

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.

Balanced by provider All models

Model	Run 1	Run 2	Run 3
Claude Fable 5 anthropic/claude-fable-5	Reduce scope	Reduce scope	Reduce scope
DeepSeek V4 Pro deepseek/deepseek-v4-pro	Reduce scope	Reduce scope	Reduce scope
Gemini 3.1 Pro Preview google/gemini-3.1-pro-preview	Reduce scope	Reduce scope	Reduce scope
GLM 5.2 z-ai/glm-5.2	Reduce scope	Reduce scope	Reduce scope
GPT-5.6 Sol Pro openai/gpt-5.6-sol-pro	Reduce scope	Reduce scope	Reduce scope
Grok 4.5 x-ai/grok-4.5	Reduce scope	Reduce scope	Reduce scope
Hy3 tencent/hy3	Add workers	Add workers	Add workers
Kimi K3 moonshotai/kimi-k3	Reduce scope	Reduce scope	Reduce scope
Laguna XS 2.1 poolside/laguna-xs-2.1	Delay launch	Delay launch	Add workers
Ling-2.6-flash inclusionai/ling-2.6-flash	Add workers	Add workers	Add workers
Llama 4 Maverick meta-llama/llama-4-maverick	Add workers	Add workers	Add workers
MiMo-V2.5-Pro xiaomi/mimo-v2.5-pro	Delay launch	Reduce scope	Add workers
MiniMax M3 minimax/minimax-m3	Add workers	Reduce scope	Reduce scope
Mistral Small 4 mistralai/mistral-small-2603	Add workers	Add workers	Reduce scope
Nemotron 3 Ultra nvidia/nemotron-3-ultra-550b-a55b	Reduce scope	Reduce scope	Reduce scope
Qwen3.7 Max qwen/qwen3.7-max	Reduce scope	Reduce scope	Reduce scope
Step 3.7 Flash stepfun/step-3.7-flash	Reduce scope	Reduce scope	Reduce scope
Claude Haiku 4.5 anthropic/claude-haiku-4.5	Add workers	Add workers	Add workers
Claude Opus 4.5 anthropic/claude-opus-4.5	Reduce scope	Reduce scope	Reduce scope
Claude Opus 4.6 anthropic/claude-opus-4.6	Reduce scope	Reduce scope	Reduce scope
Claude Opus 4.7 anthropic/claude-opus-4.7	Delay launch	Delay launch	Delay launch
Claude Opus 4.8 anthropic/claude-opus-4.8	Reduce scope	Reduce scope	Reduce scope
Claude Opus 4.8 (Fast) anthropic/claude-opus-4.8-fast	Delay launch	Delay launch	Delay launch
Claude Sonnet 4 anthropic/claude-sonnet-4	Add workers	Add workers	Add workers
Claude Sonnet 4.5 anthropic/claude-sonnet-4.5	Add workers	Add workers	Add workers
Claude Sonnet 4.6 anthropic/claude-sonnet-4.6	Add workers	Add workers	Add workers
Claude Sonnet 5 anthropic/claude-sonnet-5	Delay launch	Delay launch	Delay launch
DeepSeek V3 deepseek/deepseek-chat	Add workers	Reduce scope	Reduce scope
DeepSeek V3 0324 deepseek/deepseek-chat-v3-0324	Add workers	Add workers	Add workers
DeepSeek V3.1 deepseek/deepseek-chat-v3.1	Add workers	Reduce scope	Add workers
DeepSeek V3.1 Terminus deepseek/deepseek-v3.1-terminus	Add workers	Reduce scope	Reduce scope
DeepSeek V3.2 deepseek/deepseek-v3.2	Delay launch	Reduce scope	Add workers
DeepSeek V3.2 Exp deepseek/deepseek-v3.2-exp	Reduce scope	Reduce scope	Add workers
DeepSeek V4 Flash deepseek/deepseek-v4-flash	Add workers	Add workers	Delay launch
Gemini 2.5 Flash google/gemini-2.5-flash	Delay launch	Delay launch	Add workers
Gemini 2.5 Flash Lite google/gemini-2.5-flash-lite	Add workers	Add workers	Delay launch
Gemini 2.5 Pro google/gemini-2.5-pro	Reduce scope	Reduce scope	Reduce scope
Gemini 3 Flash Preview google/gemini-3-flash-preview	Reduce scope	Reduce scope	Reduce scope
Gemini 3.1 Flash Lite google/gemini-3.1-flash-lite	Reduce scope	Reduce scope	Reduce scope
Gemini 3.1 Flash Lite Preview google/gemini-3.1-flash-lite-preview	Reduce scope	Reduce scope	Reduce scope
Gemini 3.5 Flash google/gemini-3.5-flash	Reduce scope	Reduce scope	Reduce scope
Gemma 3 27B google/gemma-3-27b-it	Reduce scope	Reduce scope	Reduce scope
Gemma 4 26B A4B google/gemma-4-26b-a4b-it	Reduce scope	Reduce scope	Reduce scope
Gemma 4 31B google/gemma-4-31b-it	Reduce scope	Reduce scope	Reduce scope
GLM 4.7 z-ai/glm-4.7	Delay launch	Reduce scope	Reduce scope
GLM 4.7 Flash z-ai/glm-4.7-flash	Reduce scope	Reduce scope	Reduce scope
GLM 5 z-ai/glm-5	Reduce scope	Reduce scope	Reduce scope
GLM 5.1 z-ai/glm-5.1	Reduce scope	Reduce scope	Reduce scope
GPT-4.1 openai/gpt-4.1	Reduce scope	Reduce scope	Reduce scope
GPT-4.1 Mini openai/gpt-4.1-mini	Reduce scope	Reduce scope	Reduce scope
GPT-4.1 Nano openai/gpt-4.1-nano	Delay launch	Delay launch	Reduce scope
GPT-4o-mini openai/gpt-4o-mini	Reduce scope	Reduce scope	Reduce scope
GPT-5 openai/gpt-5	Reduce scope	Reduce scope	Reduce scope
GPT-5 Mini openai/gpt-5-mini	Reduce scope	Reduce scope	Reduce scope
GPT-5 Nano openai/gpt-5-nano	Add workers	Add workers	Add workers
GPT-5.1 openai/gpt-5.1	Delay launch	Reduce scope	Reduce scope
GPT-5.2 openai/gpt-5.2	Reduce scope	Reduce scope	Reduce scope
GPT-5.3-Codex openai/gpt-5.3-codex	Reduce scope	Reduce scope	Reduce scope
GPT-5.4 openai/gpt-5.4	Reduce scope	Reduce scope	Delay launch
GPT-5.4 Mini openai/gpt-5.4-mini	Reduce scope	Reduce scope	Reduce scope
GPT-5.4 Nano openai/gpt-5.4-nano	Add workers	Add workers	Add workers
GPT-5.5 openai/gpt-5.5	Reduce scope	Reduce scope	Reduce scope
GPT-5.6 Luna openai/gpt-5.6-luna	Reduce scope	Reduce scope	Reduce scope
GPT-5.6 Luna Pro openai/gpt-5.6-luna-pro	Reduce scope	Add workers	Reduce scope
GPT-5.6 Sol openai/gpt-5.6-sol	Reduce scope	Reduce scope	Reduce scope
GPT-5.6 Terra openai/gpt-5.6-terra	Reduce scope	Reduce scope	Reduce scope
GPT-5.6 Terra Pro openai/gpt-5.6-terra-pro	Reduce scope	Reduce scope	Reduce scope
gpt-oss-120b openai/gpt-oss-120b	Delay launch	Delay launch	Add workers
gpt-oss-20b openai/gpt-oss-20b	Reduce scope	Add workers	Add workers
Grok 4.20 x-ai/grok-4.20	Reduce scope	Reduce scope	Reduce scope
Grok 4.3 x-ai/grok-4.3	Reduce scope	Reduce scope	Reduce scope
Hy3 preview tencent/hy3-preview	Reduce scope	Reduce scope	Reduce scope
Kimi K2.5 moonshotai/kimi-k2.5	Reduce scope	Reduce scope	Delay launch
Kimi K2.6 moonshotai/kimi-k2.6	Reduce scope	Reduce scope	Reduce scope
Kimi K2.7 Code moonshotai/kimi-k2.7-code	Delay launch	Delay launch	Delay launch
Llama 3.1 8B Instruct meta-llama/llama-3.1-8b-instruct	Delay launch	Reduce scope	Add workers
Llama 3.3 70B Instruct meta-llama/llama-3.3-70b-instruct	Add workers	Add workers	Add workers
MiMo-V2.5 xiaomi/mimo-v2.5	Reduce scope	Reduce scope	Reduce scope
MiniMax M2.5 minimax/minimax-m2.5	Add workers	Delay launch	Add workers
MiniMax M2.7 minimax/minimax-m2.7	Delay launch	Delay launch	Add workers
Mistral Nemo mistralai/mistral-nemo	Reduce scope	Reduce scope	Reduce scope
Mistral Small 3.2 24B mistralai/mistral-small-3.2-24b-instruct	Reduce scope	Reduce scope	Reduce scope
Nemotron 3 Nano 30B A3B nvidia/nemotron-3-nano-30b-a3b	Delay launch	Delay launch	Delay launch
Nemotron 3 Super nvidia/nemotron-3-super-120b-a12b	Reduce scope	Reduce scope	Reduce scope
o4 Mini openai/o4-mini	Reduce scope	Reduce scope	Reduce scope
Qwen3 235B A22B Instruct 2507 qwen/qwen3-235b-a22b-2507	Reduce scope	Reduce scope	Reduce scope
Qwen3 30B A3B Instruct 2507 qwen/qwen3-30b-a3b-instruct-2507	Reduce scope	Reduce scope	Reduce scope
Qwen3 Coder Next qwen/qwen3-coder-next	Reduce scope	Reduce scope	Reduce scope
Qwen3 Next 80B A3B Instruct qwen/qwen3-next-80b-a3b-instruct	Reduce scope	Reduce scope	Reduce scope
Qwen3.5 397B A17B qwen/qwen3.5-397b-a17b	Reduce scope	Reduce scope	Reduce scope
Qwen3.5-122B-A10B qwen/qwen3.5-122b-a10b	Reduce scope	Reduce scope	Reduce scope
Qwen3.5-27B qwen/qwen3.5-27b	Reduce scope	Reduce scope	Reduce scope
Qwen3.5-35B-A3B qwen/qwen3.5-35b-a3b	Reduce scope	Reduce scope	Delay launch
Qwen3.5-9B qwen/qwen3.5-9b	Reduce scope	Reduce scope	Reduce scope
Qwen3.5-Flash qwen/qwen3.5-flash-02-23	Reduce scope	Reduce scope	Delay launch
Qwen3.6 27B qwen/qwen3.6-27b	Reduce scope	Reduce scope	Reduce scope
Qwen3.6 35B A3B qwen/qwen3.6-35b-a3b	Reduce scope	Reduce scope	Reduce scope
Qwen3.6 Flash qwen/qwen3.6-flash	Reduce scope	Reduce scope	Reduce scope
Qwen3.6 Plus qwen/qwen3.6-plus	Delay launch	Delay launch	Reduce scope
Qwen3.7 Plus qwen/qwen3.7-plus	Delay launch	Reduce scope	Delay launch

Most common answer Different answer 5% or less of the answers

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.

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. 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