evalytic 0.3.2

Visual Generation Quality Evaluation SDK

Evalytic

Evals for visual AI. Automated quality evaluation for AI-generated images and video.

Know if your AI-generated visuals are good — before your users tell you they're not.

pip install evalytic

evaly bench \
  -m flux-schnell -m flux-dev -m flux-pro \
  -p "A photorealistic cat on a windowsill" \
  -o report.html --yes

What It Does

Evalytic benchmarks AI image generation models by generating images, scoring them with VLM judges (Gemini, GPT, Claude, Ollama), and producing rich reports — all in one command.

• Model Selection — Compare Flux Schnell vs Dev vs Pro with real prompts
• Prompt Optimization — Measure how well models follow your prompts
• Regression Detection — Catch quality drops when models update
• CI/CD Quality Gate — Block deploys when image quality falls below threshold
• 7 Semantic Dimensions — visual_quality, prompt_adherence, text_rendering, input_fidelity, transformation_quality, artifact_detection, identity_preservation
• Consensus Judging — Multi-judge scoring with automatic agreement analysis

Quickstart

1. Install

pip install evalytic

2. Set API Keys

export FAL_KEY=your_fal_key          # fal.ai for image generation
export GEMINI_API_KEY=your_gemini_key  # Default judge

3. Run

# Single model benchmark
evaly bench -m flux-schnell -p "A cat sitting on a windowsill" --yes

# Compare models with HTML report
evaly bench -m flux-schnell -m flux-dev -m flux-pro \
  -p prompts.json -o report.html --review

# img2img benchmark
evaly bench -m flux-kontext -m seedream-edit -m reve-edit \
  -p prompts.json --input product.jpg --yes

# Score an existing image
evaly eval --image output.png --prompt "A sunset over mountains"

# CI/CD quality gate
evaly gate --report report.json --threshold 3.5

CLI Commands

Command	Description
evaly bench	Generate, score, and report in one command
evaly eval	Score a single image without generation
evaly gate	CI/CD quality gate with pass/fail exit codes

Judges

Any VLM that can analyze images works as a judge:

evaly bench -m flux-schnell -p "A cat" -j gemini-2.5-flash        # Default
evaly bench -m flux-schnell -p "A cat" -j gemini-2.5-pro           # Gemini Pro
evaly bench -m flux-schnell -p "A cat" -j openai/gpt-5.2           # OpenAI
evaly bench -m flux-schnell -p "A cat" -j anthropic/claude-sonnet-4-6  # Anthropic
evaly bench -m flux-schnell -p "A cat" -j ollama/qwen2.5-vl:7b    # Local

Consensus Mode

Use multiple judges for more reliable scores:

evaly bench -m flux-schnell -p "A cat" \
  --judges "gemini-2.5-flash,openai/gpt-5.2"

Two judges score in parallel. If they disagree, a third breaks the tie.

Optional Extras

pip install "evalytic[metrics]"  # CLIP Score + LPIPS + ArcFace (~2GB)
pip install "evalytic[all]"      # Everything

Configuration

Create evalytic.toml in your project root:

[keys]
fal = "your_fal_key"
gemini = "your_gemini_key"

[bench]
judge = "gemini-2.5-flash"
dimensions = ["visual_quality", "prompt_adherence"]
concurrency = 4

Documentation

Full docs at docs.evalytic.ai

License

MIT