Skip to content

Installation

This guide covers all installation options for Prompt Amplifier.

Basic Installation

pip install prompt-amplifier

What you get:

  • ✅ Core PromptForge class
  • ✅ TF-IDF embedder (free, no API key)
  • ✅ In-memory vector store
  • ✅ Document loaders (TXT, CSV, JSON)

What you still need:

  • ❌ API key for expand() function (OpenAI, Google, or Anthropic)
  • ❌ PDF/DOCX support (install loaders extra)
  • ❌ Semantic embeddings (install embeddings-local extra)

Installation Options

pip install prompt-amplifier[all]

What you get: Everything! All loaders, embedders, vector stores, and generators.

Size: ~2GB (includes ML models)

Minimal + Google Gemini (Lightweight)

pip install prompt-amplifier google-generativeai

What you get:

  • Core library
  • Google Gemini generator (has free tier!)

Size: ~50MB

Best for: Quick testing, Google Colab

Production Setup

pip install prompt-amplifier[loaders,embeddings-local,vectorstore-chroma]

What you get:

  • All document loaders (PDF, DOCX, Excel, etc.)
  • Sentence Transformers (semantic embeddings)
  • ChromaDB (persistent storage)

Size: ~1GB

Feature-Specific Extras

Document Loaders

pip install prompt-amplifier[loaders]

Adds support for:

Format Library Used
PDF pypdf
DOCX python-docx
Excel openpyxl
Web pages beautifulsoup4, requests
YouTube youtube-transcript-api
RSS feeds feedparser

Embedders

pip install prompt-amplifier[embeddings-local]

Adds:

  • Sentence Transformers (semantic search)
  • BM25 (keyword search)

Models used:

  • all-MiniLM-L6-v2 (384 dimensions, fast)
  • all-mpnet-base-v2 (768 dimensions, better quality)
pip install prompt-amplifier[embeddings-openai]

Adds: OpenAI embeddings (text-embedding-3-small, text-embedding-3-large)

Requires: OPENAI_API_KEY

pip install prompt-amplifier[embeddings-google]

Adds: Google embeddings (text-embedding-004)

Requires: GOOGLE_API_KEY

pip install prompt-amplifier[embeddings-cohere]

Adds: Cohere embeddings (embed-english-v3.0)

Requires: COHERE_API_KEY

Vector Stores

pip install prompt-amplifier[vectorstore-chroma]

Use case: Local persistent storage

from prompt_amplifier.vectorstores import ChromaStore

store = ChromaStore(
    collection_name="my_docs",
    persist_directory="./chroma_db"
)

pip install prompt-amplifier[vectorstore-faiss]

Use case: High-performance similarity search

from prompt_amplifier.vectorstores import FAISSStore

store = FAISSStore(index_path="./faiss_index")

LLM Generators

pip install prompt-amplifier[generators-openai]

Requires: OPENAI_API_KEY

from prompt_amplifier.generators import OpenAIGenerator
forge = PromptForge(generator=OpenAIGenerator(model="gpt-4o-mini"))

pip install prompt-amplifier[generators-anthropic]

Requires: ANTHROPIC_API_KEY

from prompt_amplifier.generators import AnthropicGenerator
forge = PromptForge(generator=AnthropicGenerator())

pip install prompt-amplifier[generators-google]
# or just: pip install google-generativeai

Requires: GOOGLE_API_KEY (free tier available!)

from prompt_amplifier.generators import GoogleGenerator
forge = PromptForge(generator=GoogleGenerator())

pip install prompt-amplifier[generators-ollama]

Requires: Ollama running locally

from prompt_amplifier.generators import OllamaGenerator
forge = PromptForge(generator=OllamaGenerator(model="llama3.2"))

Verify Installation

Run this to verify everything works:

# Test basic import
from prompt_amplifier import PromptForge
print("✅ Core library installed")

# Test with sample data
forge = PromptForge()
forge.add_texts(["Test document 1", "Test document 2"])
print(f"✅ Added {forge.chunk_count} chunks")

# Test search (no API key needed)
results = forge.search("test")
print(f"✅ Search works! Found {len(results.results)} results")

# Check available extras
try:
    from prompt_amplifier.embedders import SentenceTransformerEmbedder
    print("✅ Sentence Transformers available")
except ImportError:
    print("❌ Install with: pip install prompt-amplifier[embeddings-local]")

try:
    from prompt_amplifier.vectorstores import ChromaStore
    print("✅ ChromaDB available")
except ImportError:
    print("❌ Install with: pip install prompt-amplifier[vectorstore-chroma]")

try:
    from prompt_amplifier.generators import GoogleGenerator
    print("✅ Google generator available")
except ImportError:
    print("❌ Install with: pip install google-generativeai")

Expected Output: 

✅ Core library installed
✅ Added 2 chunks
✅ Search works! Found 2 results
✅ Sentence Transformers available (or install message)
✅ ChromaDB available (or install message)
✅ Google generator available (or install message)

Google Colab Quick Setup

For Google Colab, run these cells:

Cell 1: Install 

!pip install prompt-amplifier google-generativeai -q
print("✅ Installed!")

Cell 2: Set API Key 

import os
os.environ["GOOGLE_API_KEY"] = "your-key-from-aistudio.google.com"
print("✅ API key set!")

Cell 3: Test 

from prompt_amplifier import PromptForge
from prompt_amplifier.generators import GoogleGenerator

forge = PromptForge(generator=GoogleGenerator())
forge.add_texts(["Hello world", "Testing prompt amplifier"])

result = forge.expand("test the system")
print(result.prompt)

Troubleshooting

"No module named 'prompt_amplifier'"

# Make sure you installed it
pip install prompt-amplifier

# If using virtual environment, activate it first
source venv/bin/activate  # Linux/Mac
# or
.\venv\Scripts\activate   # Windows

"ImportError: openai is required"

pip install openai
# or install the extra
pip install prompt-amplifier[generators-openai]

"ImportError: sentence_transformers is required"

pip install sentence-transformers
# or install the extra
pip install prompt-amplifier[embeddings-local]

"No space left on device" (Colab)

Sentence Transformers downloads ~400MB models. In Colab:

# Use smaller model
from prompt_amplifier.embedders import SentenceTransformerEmbedder
embedder = SentenceTransformerEmbedder(model="all-MiniLM-L6-v2")  # Smallest

Next Steps