Introduction

PromptForge Studio V2 is a "Prompt-as-a-Service" platform designed for high-performance AI integration.

Prerequisites

Before you can execute prompts programmatically, you must complete two simple steps:

Generate an API Key: Visit the Settings sidebar in PromptForge Studio to generate your secret API key.
Save a Prompt: Create and save a prompt in the Studio. Each saved prompt version generates a unique Version ID which you'll use for execution.

Core Concepts

Dynamic Variables

Any text wrapped in {{variable_name}} in your Studio prompt is automatically treated as a dynamic template. These are substituted at runtime via the SDK or API.

Model Routing

PromptForge automatically routes your requests to the best-performing model (Flash vs Pro) based on complexity and prompt length, ensuring optimal cost-performance balance.

Semantic Caching

Our proprietary engine checks for semantically similar requests. If a match is found, the cached result is returned in sub-50ms, significantly reducing latency and costs.

Enterprise Safety

Built-in PII detection and prompt injection filters ensure that your LLM integrations are secure and compliant with enterprise standards.

Troubleshooting

401 Unauthorized

Ensure your x-api-key is correct and currently active in your dashboard settings.

404 Version Not Found

Ensure the version_id exists in your workspace and hasn't been deleted.