Unleash the Power of LLMs with Just-Prompt: Your Unified Interface to AI Giants

Tired of juggling multiple APIs to access the best Large Language Models (LLMs)? Just-Prompt offers a streamlined solution, acting as a Model Control Protocol (MCP) server to unify access to top providers like OpenAI, Gemini, Anthropic, Groq, DeepSeek, and Ollama. This article delves into how Just-Prompt simplifies LLM integration, boosts productivity, and helps you make the most of cutting-edge AI technology.

Why Choose Just-Prompt for Your LLM Needs?

• Unified API: Access multiple LLM providers through a single, consistent interface, simplifying your workflow.
• Parallel Processing: Run prompts across multiple models simultaneously and compare their outputs for optimal results.
• Easy Model Management: List available providers and models with simple commands, staying up-to-date with LLM capabilities.
• Flexible Input Options: Send prompts directly as text or load them from files, catering to different use cases.
• Automated Output Handling: Save model responses to markdown files for easy analysis and documentation.

Turn LLM Complexity into Simplicity: Essential Tools at Your Fingertips

Just-Prompt provides a range of powerful tools to streamline your LLM interactions:

• prompt: Send text prompts to multiple LLMs, specifying models or using defaults.

  • text: The prompt you want the models to process.
  • models_prefixed_by_provider (optional): A list of models to use, prefixed with their provider (e.g., openai:gpt-4o, a:claude-3-haiku).

• prompt_from_file: Load prompts from a file and send them to multiple models.

  • file: The path to the file containing your prompt.
  • models_prefixed_by_provider (optional): A list of target models.

• prompt_from_file_to_file: Process prompts from a file and save the responses to separate markdown files.

  • file: The path to the prompt file.
  • models_prefixed_by_provider (optional): The models to use.
  • output_dir (optional): Directory for saving responses (defaults to ".").

• ceo_and_board: Simulate a decision-making scenario with "board member" models and a "CEO" model that synthesizes their input. This tool leverages the power of LLM orchestration to tackle complex challenges.

  • file: Path to the file containing the prompt describing the scenario.
  • models_prefixed_by_provider (optional): List of models acting as board members.
  • output_dir (optional): Directory for saving responses and the CEO's decision (defaults to ".").
  • ceo_model (optional): Model for the CEO role (defaults to "openai:o3").

• list_providers: Display available LLM providers supported by Just-Prompt.

• list_models: List all models available for a specific provider.

  • provider: The provider to query (e.g., 'openai' or 'o').

Provider Prefixes: Quick Access to Your Favorite Models

Just-Prompt uses short prefixes for each provider to make model selection faster and more convenient:

• OpenAI: o or openai (e.g., o:gpt-4o-mini)
• Anthropic: a or anthropic (e.g., a:claude-3-5-haiku)
• Google Gemini: g or gemini (e.g., g:gemini-2.5-pro-exp-03-25)
• Groq: q or groq (e.g., q:llama-3.1-70b-versatile)
• DeepSeek: d or deepseek (e.g., d:deepseek-coder)
• Ollama: l or ollama (e.g., l:llama3.1)

Using these prefixes, you can easily specify the desired model for each operation.

Installing and Configuring Just-Prompt

1. Clone the Repository: git clone https://github.com/yourusername/just-prompt.git; cd just-prompt

2. Install Dependencies: uv sync

3. Configure API Keys: Create a .env file with your API keys (copy from .env.sample):

   OPENAI_API_KEY=your_openai_api_key_here
   ANTHROPIC_API_KEY=your_anthropic_api_key_here
   GEMINI_API_KEY=your_gemini_api_key_here
   GROQ_API_KEY=your_groq_api_key_here
   DEEPSEEK_API_KEY=your_deepseek_api_key_here
   OLLAMA_HOST=http://localhost:11434
   

Optimize LLM Performance: Reasoning Effort, Thinking Tokens, and Thinking Budget

Just-Prompt allows you to fine-tune the reasoning capabilities of certain models:

• OpenAI's o-Series: Control reasoning effort with suffixes :low, :medium, or :high (e.g., openai:o4-mini:low).

• Anthropic's Claude-3-Sonnet: Leverage thinking tokens by adding a suffix like :1k, :4k, or :8000 to specify the token budget (e.g., anthropic:claude-3-7-sonnet-20250219:4k).

• Google's Gemini-2.5-Flash: Utilize thinking budget with suffixes like :1k, :4k, or :8000 (e.g., gemini:gemini-2.5-flash-preview-04-17:4k).

Getting Started with Claude Code Integration

Just-Prompt seamlessly integrates with Claude Code, a powerful development environment. Use the mcp add command to register Just-Prompt. Here are a few examples:

• Basic Integration:

  claude mcp add just-prompt "$(pbpaste)"
  # JSON to copy:
  # { "command": "uv", "args": ["--directory", ".", "run", "just-prompt"] }
  

• With Default Models:

  claude mcp add just-prompt -s project \
  -- \
  uv --directory. \
  run just-prompt --default-models " openai:gpt-4o "
  

Elevate Your LLM Workflow Today

Just-Prompt empowers you to harness the collective intelligence of multiple LLMs with ease. By providing a unified interface and advanced control over model behavior, it streamlines your AI development process and unlocks new possibilities for innovation. Start leveraging the power of Just-Prompt today!