# Benchmarks

> Samsa's benchmarks rank current AI models and tools on one task each, with the method, the costs and the raw scores published for every run.

Human version: https://samsa.ai/benchmarks/
Language: English
Machine index: https://samsa.ai/llms.txt

---
Samsa's benchmarks measure current AI models and tools on one practical task each, under a published protocol: the same prompt and the same inputs for every participant, a score from published metrics, and the cost and time of every run shown beside the score. The raw numbers of each run are published, and new models are added as they ship.

---

- [Vectorization benchmark: which AI turns an illustration into an editable SVG?](https://samsa.ai/benchmarks/vectorization/) — Updated 2026-10-08. Vectorization benchmark: Claude, GPT, Gemini, Recraft and Quiver AI scored on turning an illustration into a layered, editable SVG, with cost and time.