Benchmarks

Published results from running the same prompts on different AI models. Each one opens its own reader.

Writing

Writing comparisons

Several models

Czech poetry, Nanynka rewrites, Wodehouse, Mary Had a Little Lamb in ten poets' styles, and Seifert translations.

Open reader →

Drawing and images

Bicycle Test, October 2026

13 Oct 2026 · Claude Haiku 5.5, Opus 5.5, Sonnet 5.5; Muse Spark 1.3, Wonder 1.0, Wonder Full 1.0

Each model was asked to make an SVG of a bicycle.

Open reader →

Bicycle Test, September 2026

4 Sep 2026 · Claude Fable 5.1, Gemini 3.8 Flash, GPT-6 Astra, GLM 5.3, GLM 5.3 Flash

Each model was asked to make an SVG of a bicycle.

Open reader →

Bicycle Test, Google models

12 Jul 2026 · Google models

Google models were each asked to make an SVG of a bicycle.

Open reader →

Video

Gemma 4 video understanding

June 2026 · Gemma 4

Two videos were analysed: a screencast and a meeting clip.

Open reader →

Speech

Czech speech transcription

Date not recorded · Deepgram Nova-2, ElevenLabs Scribe, Gemini 2.0 Experimental, Rev.ai, Whisper

Five speech-to-text services transcribed the same Czech recording.

Open reader →

General vibe check

Model vibe run

July 2026 · Gemini 3 Flash Preview

71 prompts were run on one model.

Open reader →