Directory / DSPy
DSPy
Verified
Stanford framework for programming language-model behavior.
DSPy is an open-source framework for compiling LM programs from signatures, modules, and optimizers instead of ad-hoc prompts. Teams use it as an orchestration and intent layer when the factory must change prompts against a metric.
- Model support
- Multi-model
- Deployment
- Self-hosted
- Open source
- Yes
- License
- MIT
- Pricing model
- Free
- Added / updated
- 19 Aug 2026 / 19 Aug 2026
Assessment
DSPy handles prompt and model optimization against a metric. Pull-request creation and worker compute sit outside the framework.
Strengths
- Programmatic prompts, optimizers, and public MIT source.
- Model-agnostic interfaces.
- Metrics are part of the optimization loop.
Limitations
- Research-oriented learning curve.
- No execution sandbox.
- Optimizers can overfit small evaluation sets.