priors
Support

Settings

Theme

Explore themes

Mode

Library

Looks borrowed from benchmark sites and products. Themes with a light and a dark version follow your mode.

Bring your own theme

Upload a Priors theme file, or have your agent make one: copy the prompt, tell it the look you want, and upload what it returns. With the Priors skill installed it already knows the format.

Custom themes stay in this browser. They travel inside shared chart links and a page's theme, so others see them as you do.

For the builder, Explore and model pages on this device. Published pages keep the theme their owner picked.

Mode

Benchmarks · agentic

CUA-bench

Computer-use agents playing commercial video games through keyboard and mouse input, with half the titles held out.

8 models · 8 results · updated 2026-10-13 · vals.ai ↗

Leaderboard

Chart settings
8 of 8 selected
GPT-6 Astra max

Score vs subscription cost per task

CUA-bench · measured cost per task · subscription-priced where plan data exists (Real API Pricing, full allowance used)

Highlight models

Click a point on the chart to highlight that exact effort; click it again to remove it.

3 more in settings
Chart settings
8 of 8 selected

More from this benchmark

1 published measures
Time per task9 results · lower is better

Highlight models

Click a point on the chart to highlight that exact effort; click it again to remove it.

3 more in settings
Chart settings
8 of 8 selected

All results

8 rows
#ModelEffortCUA-benchSubscription $ / taskAPI $ / taskTimeHarness
1GPT-6 Astramax19.2%$156$182132scodex
2Opus 5.5max14.0%$2.85$18815sclaude-code
3Fable 5.1max13.2%$50$32437sclaude-code
4Opus 5max9.0%$4.03$23225sclaude-code
5GPT-5.6 Solmax8.3%$11$20719scodex
6Muse Spark 1.3max5.8%$72$14336smuse-code
7Gemini 4 Argonhigh4.8%$69$19471santigravity
8Gemini 3.8 Flashhigh4.2%$11$28037sgoogle-computer-use