Skip to main content

​
TL;DR

  • Multi-Stage Optimizers: Both MIPRO and GEPA now support multi-stage pipeline optimization
  • GEPA Algorithm: Genetic Evolution for Prompt Optimization now available
  • Expanded Model Support: Google Gemini and OpenAI models added
  • Enhanced SDK validation with comprehensive error checking

​
Multi-Stage Optimizers & Expanded Model Support

Both prompt optimization algorithms now support multi-stage pipeline optimization for complex workflows with multiple processing stages.

​
MIPRO Multi-Stage

  • Generates per-stage instruction proposals with automatic stage detection via LCS (Longest Common Subsequence) matching
  • Each stage gets stage-specific meta-prompts including pipeline overview, stage role, and baseline performance
  • Supports per-module configuration with max_instruction_slots and max_demo_slots for fine-grained control

​
GEPA Multi-Stage

  • Uses module-aware evolution where each pipeline module gets its own gene
  • Mutations target specific modules, uniform crossover combines parent genes per module
  • Aggregated scoring sums module lengths for Pareto optimization
  • Supports per-module max_instruction_slots, max_tokens, and allowed_tools configuration

​
Configuration

Both algorithms support pipeline_modules metadata in initial prompts and module-specific settings in their respective config sections (prompt_learning.gepa.modules and prompt_learning.mipro.modules).

​
GEPA: Genetic Evolution for Prompt Optimization

GEPA algorithm is now available for prompt optimization jobs.

​
Key Features

  • Evolutionary Algorithms: Uses mutation, crossover, and selection to optimize prompts across multiple generations
  • Multi-Objective Optimization: Maintains a Pareto front balancing accuracy, token count, and task-specific metrics
  • Prompt ID-Based URLs: Prompt transformations use versioned URLs (/v1/{prompt_version_id}/chat/completions) for better traceability
  • Validation Scoring: Job results distinguish between prompt_best_train_score and prompt_best_validation_score

​
Integration Testing

Added comprehensive integration tests for GEPA training workflows with Banking77 task app.

​
Gemini Model Support

Added comprehensive support for Google Gemini models as policy models for both GEPA and MIPRO algorithms.
  • Supported Models: gemini-2.5-pro (≤200k tokens), gemini-2.5-pro-gt200k (>200k tokens), gemini-2.5-flash, and gemini-2.5-flash-lite
  • Provider Integration: Full SDK validation and backend support for provider = "google" with automatic pricing calculation
  • Example Configs: Added example configurations demonstrating Gemini usage

​
OpenAI Model Support

Expanded OpenAI model support for prompt optimization with comprehensive coverage of latest models.
  • Supported Models: gpt-4o, gpt-4o-mini, gpt-4.1, gpt-4.1-mini, gpt-4.1-nano, gpt-5, gpt-5-mini, and gpt-5-nano
  • Model Validation: SDK-side validation with clear error messages for unsupported models
  • Explicit Rejection: gpt-5-pro rejected due to high cost (15/15/120 per 1M tokens)
  • Provider Prefix Support: Models can be specified with or without provider prefix

​
SDK Validation Enhancements

Improved config validation with comprehensive error checking before sending to backend.
  • Multi-Stage Validation: Validates that pipeline_modules match module configs, checks for missing or extra modules
  • Model Validation: Provider-aware model validation with detailed error messages listing supported models
  • Nano Model Restrictions: Clear validation that nano models are allowed for policy models but rejected for mutation/meta models

​
Documentation

  • Multi-Stage Pipeline Guide: Updated documentation with examples and configuration details
  • Model Support Reference: Complete documentation of supported models for each provider
  • GEPA Guide: Complete documentation with quick start, configuration examples, and troubleshooting
  • Example Configurations: Added example configs demonstrating multi-stage optimization

​
Use Cases

  • Complex Pipelines: Optimize multi-stage workflows with stage-specific prompt improvements
  • Evolutionary Optimization: Use GEPA for population-based prompt optimization
  • Multi-Provider Support: Choose from OpenAI, Google, or Groq models based on your needs
  • Cost Optimization: Use nano models for policy evaluation while using larger models for generation