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

Anthropic · released 2026-09-28

Sonnet 5.5

Measured on 164 benchmarks across 5 reasoning efforts.

54.3

#5 of 61 models · best at max

Reasoning efforts in this ranking

max54.3#9 of 139
xhigh49.6#18 of 139
high41.1#33 of 139
medium30.2#60 of 139
low25.3#83 of 139

Strengths

  • Sonnet 5.5
  • Gemini 4 Carbon (#1)
  • Field median
bestField medianFrontierCode 2.0-…Terminal-Bench v4…AA-Briefcase v1.1AutomationBench-AATerminal-Bench-Sc…HLE-DiamondCritPt v2 Hard Su…AA-Omniscience v1…AA-Omniscience v1…GDP.pdf
  1. FrontierCode 2.0-AA15%41.7%#4 of 61
  2. Terminal-Bench v4-AA10%63.6%#3 of 61
  3. AA-Briefcase v1.115%1814#1 of 61
  4. AutomationBench-AA5%71.8%#5 of 61
  5. Terminal-Bench-Science-AA15%53.3%#6 of 61
  6. HLE-Diamond10%53.3%#7 of 61
  7. CritPt v2 Hard Subset10%56.0%#8 of 61
  8. AA-Omniscience v1.1 Accuracy10%51.5%#10 of 61
  9. AA-Omniscience v1.1 Non-hallucination5%56.7%#33 of 61
  10. GDP.pdf5%25.8%#12 of 61

Bars show where Sonnet 5.5 sits among the 61 ranked models on each benchmark (right is best); the tick is the field median.

Cost and speed

against 61 ranked models

Cost per task

#16 cheapest of 41

$0.15at max

Time per task

#32 quickest of 40

21 minat max

Output speed

#19 fastest of 52

139 tok/sat max

Time to first token

#51 quickest of 52

7.6 minat max

Output tokens per task

#44 fewest of 44

285kat max

Token price

$0.196 per 1M tokens at API prices$0.0030 on Claude Max 20x

97.3% cache read · 0.04% fresh input · 2.20% cache write · 0.43% output

Measured from T3 Code usage in Claude Code, Oct 27 to Nov 2, high confidence. One app’s users, not everyone’s. $0.197 at the standard 97 / 2.5 / 0.5 mix.

This ranking’s benchmarks

10 of 10 measured

agentic

AutomationBench-AA

241 results
  1. #6max of 24171.8%
  2. #21xhigh of 24165.5%
  3. #46high of 24159.4%
  4. #70medium of 24154.9%
  5. #93low of 24149.4%

Leader: Gemini 4 Carbon · 77.8%

agentic

AA-Briefcase v1.1

247 results
  1. #1max of 2471814
  2. #4xhigh of 2471744
  3. #13high of 2471635
  4. #54medium of 2471440
  5. #88low of 2471268

Leader: this model · 1814

reasoning

CritPt v2 Hard Subset

100 results
  1. #23max of 10056.0%
  2. #26xhigh of 10054.0%
  3. #33high of 10047.5%
  4. #60medium of 10028.5%
  5. #71low of 10022.5%

Leader: GPT-6 Astra · 71.5%

coding

FrontierCode 2.0-AA

97 results
  1. #5max of 9741.7%
  2. #19xhigh of 9735.6%
  3. #35high of 9725.7%
  4. #71medium of 9711.0%
  5. #77low of 977.1%

Leader: Opus 5.5 · 46.4%

economics

GDP.pdf

240 results
  1. #34max of 24025.8%
  2. #38high of 24025.2%
  3. #40xhigh of 24024.6%
  4. #72medium of 24020.2%
  5. #103low of 24016.0%

Leader: Muse Wonder Full · 32.6%

reasoning

HLE-Diamond

142 results
  1. #15max of 14253.3%
  2. #26xhigh of 14244.1%
  3. #34high of 14233.8%
  4. #53medium of 14223.8%
  5. #61low of 14220.6%

Leader: Gemini 4 Carbon · 69.1%

knowledge

AA-Omniscience v1.1 Accuracy

111 results
  1. #22max of 11151.5%
  2. #23xhigh of 11150.9%
  3. #25high of 11150.3%
  4. #36low of 11146.0%
  5. #37medium of 11145.8%

Leader: Fable 5.1 · 68.4%

knowledge

AA-Omniscience v1.1 Non-hallucination

111 results
  1. #49max of 11156.7%
  2. #54low of 11152.3%
  3. #61medium of 11150.0%
  4. #71xhigh of 11138.3%
  5. #73high of 11136.4%

Leader: Gemini 4 Carbon · 95.8%

reasoning

Terminal-Bench-Science-AA

115 results
  1. #11max of 11553.3%
  2. #14xhigh of 11552.4%
  3. #27high of 11534.3%
  4. #42medium of 11516.2%
  5. #51low of 11510.5%

Leader: Gemini 4 Carbon · 64.8%

coding

Terminal-Bench v4-AA

238 results
  1. #3max of 23863.6%
  2. #12xhigh of 23857.1%
  3. #31high of 23843.9%
  4. #54medium of 23829.8%
  5. #74low of 23820.7%

Leader: Kimi K3.1 · 67.2%

All 164 benchmarks

show 154 more