DevXTools
DEVELOPER TOOL

Prompt Template Playground

Render reusable prompt templates with automatically discovered variables and local token estimates.

Runs locallyRuns locally in your browser. Your data is not uploaded.
Prompt template workspace

Template and variables

Use {{variableName}}: 1–64 ASCII letters, digits or underscores, starting with a letter or underscore. No spaces or expressions. Single braces are literal. Template: 16 KiB UTF-8; up to 32 variables and 128 occurrences.

Each value: 8 KiB UTF-8. Empty values preserve their placeholder; whitespace-only values are literal. Values are substituted once, never interpreted. Removed variables are discarded.

Rendered prompt

Output: 64 KiB UTF-8. Token counts reuse the Token Counter’s local UTF-8 weighted estimate, including unresolved placeholders. Not exact or a guaranteed bound. No verified local GPT-6.1 encoding is bundled. This is a text heuristic, not the provider tokenizer. API wrappers, tools, media and output tokens are excluded.

Ready. Variables are discovered as you type.

Enable JavaScript to render locally.

Variables, not expressions.

Use double braces around an ASCII identifier, such as {{user_name}}. Each unique name gets one field. Single braces are literal; malformed double-brace placeholders are rejected. No functions, conditions, loops, imports or recursive expansion are supported.

Missing values stay visible.

An absent or empty value preserves its placeholder and appears in the unresolved list. Whitespace values remain literal. Counts include preserved placeholders. Template edits discard fields that no longer exist; reset and leaving the page clear content.

Local estimates with explicit limits.

Template: 16 KiB UTF-8, 32 unique variables and 128 occurrences. Each value: 8 KiB; rendered output: 64 KiB. Characters count Unicode code points and words are whitespace-delimited. Token estimates reuse the Token Counter heuristic; they are not exact or guaranteed bounds. Copy sends text only to your clipboard.

On the same workbench.

Local

LLM Token Counter

Estimate prompt tokens locally with clear accuracy labels and model-specific tokenizer disclosures.

AI & LLMOpen tool
Local

Context Window Calculator

Check input, output reserves and remaining model context locally.

AI & LLMOpen tool
Local

LLM API Request Builder

Build OpenAI Responses, Anthropic Messages and Gemini Interactions examples locally, without sending requests or collecting keys.

AI & LLMOpen tool

Find a tool or category

Tools & categories12 results
↑↓ to navigate↵ to openesc to close Search stays local