Installation

Table of Contents

• Installation Methods
  • Download Binary
  • Install Using Go
  • Build from Source
• Authentication
  • Gemini Provider
  • OpenAI Provider

Installation Methods

There are several ways to install prompt2json:

Download Binary

Download pre-built binaries from the GitHub Releases page for the latest version.

Choose the appropriate binary for your platform and add it to your PATH.

Install Using Go

Install directly from the Go toolchain:

go install github.com/UnitVectorY-Labs/prompt2json@latest

Build from Source

Build the application from source code:

git clone https://github.com/UnitVectorY-Labs/prompt2json.git
cd prompt2json
go build -o prompt2json

Authentication

Gemini Provider

prompt2json requires Google Cloud credentials to access Gemini models when using the Gemini provider (--provider gemini).

You will be charged for usage of Gemini models according to Google Cloud’s pricing.

Authenticate locally:

gcloud auth application-default login

Or use a service account:

export GOOGLE_APPLICATION_CREDENTIALS=/path/to/key.json

Set your project (can be specified with --project flag as well):

export GOOGLE_CLOUD_PROJECT=your-project-id

OpenAI Provider

You will be charged for usage of OpenAI models according to OpenAI’s pricing.

When using the OpenAI provider (--provider openai), an API key is required when using the default OpenAI URL:

export OPENAI_API_KEY=your-api-key

Or provide it directly via the --api-key flag.

When using --url to specify a custom endpoint (such as a local Ollama server), the API key is optional.