polyvise / model lab
Model lab

Compare models on the same prompt

Send one prompt to every configured model and compare the answers side by side, along with what each one cost and how long it took. Models that fail or have no API key stay in the table rather than disappearing from it.

Roster6 providers · 12 models
ModelProviderDefault slotSpeedTierState
openai/gpt-5.5openaideepbalancedpremiumready
openai/gpt-5.6-solopenaideliberatepremiumready
openai/gpt-4.1openaibalancedstandardready
openai/gpt-4o-miniopenaifasteconomyready
anthropic/claude-opus-4.8anthropicdeliberatepremiumready
anthropic/claude-sonnet-4.6anthropicbalancedstandardready
google/gemini-3.1-flash-litegooglefasteconomyready
google/gemini-2.5-flashgooglequickfasteconomyready
deepseek/deepseek-v4-prodeepseekdeliberatestandardready
deepseek/deepseek-v4-flashdeepseekfasteconomyready
x-ai/grok-4.5xaibalancedpremiumready
z-ai/glm-5.2zaibalancedstandardready
2 of 3 models selected

Every selected model gets the identical prompt. Models that fail stay in the results with the reason rather than disappearing from them.