/MODELS [X 44.8, Y 18.2]
Seven models, one adapter surface.
Each model has a native format · ChatML, XML tags, Gemini native, Llama template, Mistral native. Prompt Studio adapts your prompt into every one and reports the cross-model compatibility matrix in one call.
[X 44.8, Y 18.2]
/S.01 Supported
ModelProviderContextFormat
GPT-4oOpenAI128KChatML
Claude 3.5 SonnetAnthropic200KXML Tags
Gemini 1.5 ProGoogle1MGemini Native
GPT-3.5 TurboOpenAI16KChatML
Llama 3.1 70BMeta128KLlama Template
Mistral LargeMistral AI32KMistral Native
DeepSeek-V3DeepSeek64KChatML
/S.02 Scoring dimensions
Seven axes, one grade.
/S.02.01ClaritySentence structure, optimal word count (~40–80 words ideal)
/S.02.02SpecificityPresence of clear action verbs
/S.02.03Context richnessRole definition, background, few-shot examples
/S.02.04Format specExplicit output format (JSON, markdown, bullet list, etc.)
/S.02.05Mode alignmentVocabulary match for TECHNICAL / CREATIVE / SYSTEM mode
/S.02.06Token efficiencyLength relative to task complexity
/S.02.07ConstraintsBoundaries, guardrails, and scope limiters