priorsBuilder
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

Update notices

A note when a benchmark one of your indexes uses gets a new version, with a one-click upgrade.

Google · released 2026-11-12

Gemini 4 Hydrogen

Measured on 45 benchmarks across 4 reasoning efforts.

53.9

#7 of 72 models · best at xhigh

Reasoning efforts in this ranking

xhigh53.9#15 of 163
high50.8#21 of 163
medium41.2#41 of 163
low26.9#92 of 163

Strengths

  • Gemini 4 Hydrogen
  • Fable 5.2 (#1)
  • Field median
bestField medianFrontierCode 2.0-AATerminal-Benchv4-AAAutomationBench-AAAA-Briefcase v1.1GDP.pdfTerminal-Bench-Science-AACritPt v2 HardSubsetAA-Omnisciencev1.1 AccuracyHLE-DiamondAA-Omnisciencev1.1Non-hallucination
  1. FrontierCode 2.0-AA15%39.0%#8 of 72
  2. Terminal-Bench v4-AA10%53.5%#12 of 72
  3. AutomationBench-AA5%76.1%#4 of 72
  4. AA-Briefcase v1.115%1567#17 of 72
  5. GDP.pdf5%27.6%#8 of 72
  6. Terminal-Bench-Science-AA15%49.5%#9 of 72
  7. CritPt v2 Hard Subset10%66.0%#8 of 72
  8. AA-Omniscience v1.1 Accuracy10%50.5%#12 of 72
  9. HLE-Diamond10%58.8%#5 of 72
  10. AA-Omniscience v1.1 Non-hallucination5%91.5%#2 of 72

Bars show where Gemini 4 Hydrogen sits among the 72 ranked models on each benchmark (right is best); the tick is the field median.

By category

  • Gemini 4 Hydrogen
  • Fable 5.2 (#1)
  • Field median
p25p50p75topField medianCodingAgents & computeruseProfessional workScience & researchReasoning & puzzlesGames & strategyKnowledge &accuracyWriting & design
  1. Coding7 benchmarks80%#5 of 72
  2. Agents & computer use16 benchmarks81%#4 of 72
  3. Professional work9 benchmarks83%#6 of 72
  4. Science & research3 benchmarks89%#2 of 71
  5. Reasoning & puzzles2 benchmarks90%#3 of 69
  6. Games & strategy2 benchmarks86%#4 of 45
  7. Knowledge & accuracy2 benchmarks93%#2 of 68
  8. Writing & design2 benchmarks55%#17 of 44

Across 43 results on every current benchmark that measured Gemini 4 Hydrogen, not only this ranking’s. The percentage is its average standing among the 72 models in this ranking that took the same benchmarks: 100% beat them all, 0% trailed them all. A benchmark in two categories counts in both; a category needs two benchmarks.

Cost and speed

against 72 ranked models

Cost per task

#50 cheapest of 55

$2.3at xhigh

Time per task

#21 quickest of 54

14 minat xhigh

Output speed

#10 fastest of 61

190 tok/sat xhigh

Time to first token

#43 quickest of 61

38sat xhigh

Output tokens per task

#51 fewest of 60

154kat xhigh

Time per task: Artificial Analysis · computed, not timed: output tokens per Intelligence Index task divided by output speed, without tool time. Method changed Nov 4, 2026.

Token price

$0.319 per 1M tokens at API prices$0.227 on Google AI Pro (assumed same as Argon)

85.0% cache read · 14.5% fresh input · 0.00% cache write · 0.50% output

Assumed mix: no measured T3 Code usage for this model.

This ranking’s benchmarks

10 of 10 measured

Cost per task is what a subscriber pays: the model’s plan price, a share of API price. Models no plan sells are at API price.

Agents & computer use

AutomationBench-AA

259 results
  1. #5xhigh of 25976.1%
  2. #6high of 25974.4%
  3. #10medium of 25972.0%
  4. #50low of 25960.5%

Leader: Gemini 4 Carbon · 77.8%

Professional work

AA-Briefcase v1.1

264 results
  1. #32xhigh of 2641567
  2. #51high of 2641497
  3. #94medium of 2641296
  4. #163low of 264947

Leader: Sonnet 5.5 · 1816

Science & research

CritPt v2 Hard Subset

133 results
  1. #18xhigh of 13366.0%
  2. #24high of 13361.5%
  3. #26medium of 13361.0%
  4. #49low of 13348.0%

Leader: Fable 5.2 · 76.5%

Coding

FrontierCode 2.0-AA

131 results
  1. #15xhigh of 13139.0%
  2. #23high of 13136.3%
  3. #34medium of 13131.5%
  4. #96low of 1319.3%

Leader: Fable 5.2 · 49.1%

Professional work

GDP.pdf

257 results
  1. #1medium of 25732.8%
  2. #13high of 25730.0%
  3. #22low of 25728.2%
  4. #26xhigh of 25727.6%

Leader: this model · 32.8%

Knowledge & accuracy

HLE-Diamond

167 results
  1. #16xhigh of 16758.8%
  2. #21high of 16754.9%
  3. #28medium of 16747.7%
  4. #48low of 16732.2%

Leader: Fable 5.2 · 80.2%

Knowledge & accuracy

AA-Omniscience v1.1 Accuracy

140 results
  1. #32xhigh of 14050.5%
  2. #39medium of 14048.3%
  3. #42high of 14048.2%
  4. #48low of 14046.2%

Leader: Fable 5.2 · 79.1%

Knowledge & accuracy

AA-Omniscience v1.1 Non-hallucination

140 results
  1. #3xhigh of 14091.5%
  2. #4high of 14091.3%
  3. #6medium of 14087.5%
  4. #13low of 14081.2%

Leader: Gemini 4 Carbon · 95.8%

Science & research

Terminal-Bench-Science-AA

142 results
  1. #21xhigh of 14249.5%
  2. #27high of 14241.4%
  3. #53medium of 14215.2%
  4. #98low of 1422.4%

Leader: Fable 5.2 · 73.3%

Agents & computer use

Terminal-Bench v4-AA

255 results
  1. #24high of 25554.0%
  2. #25xhigh of 25553.5%
  3. #67medium of 25529.3%
  4. #128low of 2556.6%

Leader: Kimi K3.1 · 67.2%

All 45 benchmarks

show 35 more