Skip to content

SteadyText Documentation

Deterministic AI for app developers and database teams. Same inputs, same outputs—across Python and Postgres.

Versioning

Releases follow yyyy.mm.dd. The current docs track the 2025.10.xx train covering both the Python SDK and the pg_steadytext extension.

Choose Your Track

Time-to-first-deterministic-result in minutes. Install the SDK, run your first steadytext.generate, and wire deterministic workflows into apps or CI.

Bring deterministic generation, embeddings, and reranking into SQL. Install the extension, connect to the daemon, and build pipelines inside Postgres.

Core Platform at a Glance

  • Daemon-backed services keep models hot and provide shared caching across SDK and SQL.
  • Deterministic execution guarantees reproducible outputs with seed control and graceful fallbacks.
  • Structured generation & reranking work the same in Python and SQL, powered by a shared model registry.
  • Vector indexing via FAISS integrates with CLI helpers and Postgres search functions.

Read the Core Platform Hub to learn how the pillars connect and when to choose each surface.

Why Teams Ship with SteadyText

  • Predictable CI – deterministic responses eliminate flaky AI-based tests.
  • Hybrid deployment – mix Python services with SQL automation without managing separate model stacks.
  • Timeboxed iteration – date-based versions and changelog callouts make upgrades deliberate.
  • Open ecosystem – models, cache backends, and integrations are swappable via configuration.

See Benchmarks for performance data and FAQ for quick answers.

What’s New

  • Twin Pillars navigation: mirrored quick starts and tutorials for Python and Postgres.
  • Expanded Postgres coverage: async workflows, reranking, and prompt registry guidance.
  • Operations consolidation: deployment, caching, and unsafe-mode docs share a single home.

Check Version History and Changelog (GitHub) for release specifics.

Need Help?

You're now ready to pick a track and explore. Happy shipping!

📦 Installation & Models

Install stable release:

# Using UV (recommended - 10-100x faster)
uv add steadytext

# Or using pip
pip install steadytext

Models

Current models (v2025.8.17+):

  • Generation (Small): Qwen3-4B-Instruct (3.9GB) - High-quality 4B parameter model (default)
  • Generation (Large): Qwen3-30B-A3B-Instruct (12GB) - Advanced 30B parameter model
  • Embeddings: Jina-v4-Text-Retrieval (1.2GB) - State-of-the-art 2048-dim embeddings (truncated to 1024)
  • Reranking: Qwen3-Reranker-4B (3.5GB) - Document reranking model

Version Stability

Each major version will use a fixed set of models only, so that only forced upgrades from pip will change the models (and the deterministic output)

🎯 Use Cases

Perfect for

  • Testing AI features: Reliable asserts that never flake
  • Deterministic CLI tooling: Consistent outputs for automation
  • Reproducible documentation: Examples that always work
  • Offline/dev/staging environments: No API keys needed
  • Semantic caching and embedding search: Fast similarity matching

Not ideal for

  • Creative or conversational tasks
  • Latest knowledge queries
  • Large-scale chatbot deployments

📋 Examples

Use SteadyText in tests or CLI tools for consistent, reproducible results:

# Testing with reliable assertions
def test_ai_function():
    result = my_ai_function("test input")
    expected = steadytext.generate("expected output for 'test input'")
    assert result == expected  # No flakes!

# CLI tools with consistent outputs
import click

@click.command()
def ai_tool(prompt):
    print(steadytext.generate(prompt))

📂 More examples →

🔍 API Overview

# Text generation
steadytext.generate(prompt: str) -> str
steadytext.generate(prompt, return_logprobs=True)
steadytext.generate(prompt, schema=MyModel)  # Structured output

# Streaming generation
steadytext.generate_iter(prompt: str)

# Document reranking
steadytext.rerank(query: str, documents: List[str]) -> List[Tuple[str, float]]

# Embeddings
steadytext.embed(text: str | List[str]) -> np.ndarray

# Model preloading
steadytext.preload_models(verbose=True)

📚 Full API Documentation →

🔧 Configuration

Control caching behavior via environment variables:

# Generation cache (default: 256 entries, 50MB)
export STEADYTEXT_GENERATION_CACHE_CAPACITY=256
export STEADYTEXT_GENERATION_CACHE_MAX_SIZE_MB=50

# Embedding cache (default: 512 entries, 100MB)
export STEADYTEXT_EMBEDDING_CACHE_CAPACITY=512
export STEADYTEXT_EMBEDDING_CACHE_MAX_SIZE_MB=100

🤝 Contributing

Contributions are welcome! See Contributing Guide for guidelines.

📄 License

  • Code: MIT
  • Models: MIT (Qwen3)

Built with ❤️ for developers tired of flaky AI tests.