Sari la conținut

Română · beta. Traducerea a fost făcută automat și este încă în curs de revizuire. Citește pagina în engleză

Benchmark-uri

Surse
8
Intrări urmărite
22.318
Ultima sincronizare
26 min
Toate sursele se sincronizează normal
  • Arena · Textacum 2 ore
  • Arena · WebDevacum 2 ore
  • Arena · Visionacum 2 ore
  • Arena · Agentacum 2 ore
  • GPQA Diamondacum 26 de minute
  • SWE-benchacum 26 de minute
  • Terminal-Benchacum 26 de minute
  • ARC-AGI-2acum 26 de minute

gpt-6-astra

95.0%

niciun interval publicat

OpenAIsept. 2026oscilație #1–#18

gpt-6.1-sol

94.2%

niciun interval publicat

OpenAIoct. 2026oscilație #1–#15

claude-opus-5-5

93.3%

niciun interval publicat

Anthropicoct. 2026oscilație #2–#11

ARC-AGI-2 ×189 din 89 modele
ModelOrganizație± CIOscilațieAdaugă la comparație
1gpt-6-astraOpenAIsept. 2026$10 / $5095.0%—#1–18
2gpt-6.1-solOpenAIoct. 2026$2 / $1094.2%—#1–15
3claude-opus-5-5Anthropicoct. 2026$4 / $2093.3%—#2–11
4gpt-5.6-solOpenAIsept. 2026$4 / $2092.5%—#6–24
5claude-opus-5Anthropicsept. 2026$5 / $2590.4%—#5–11
6claude-fable-5-1Anthropicsept. 2026$10 / $5090.0%—#2–22
7gpt-6-solOpenAIoct. 2026$2 / $1089.6%—#7–53
8claude-fable-5Anthropicsept. 2026$10 / $5089.2%—#4–21
9gemini-3.8-flashGoogle DeepMindoct. 2026$0.83 / $4.1389.2%—#2–30
10gpt-5.5OpenAIsept. 2026$5 / $3085.0%—#9–34
11gemini-3.7-flashGoogle DeepMindoct. 2026$0.75 / $3.7584.6%—#9–57
12gemini-3-deep-think-previewGoogle DeepMind——84.6%—#84–163
13gpt-5.5-proOpenAI—$30 / $18084.6%—#59–85
14gpt-5.6-terraOpenAIsept. 2026$2 / $1283.9%—#13–60
15gpt-5.4-pro-2026-03-05OpenAImar. 2026$30 / $18083.3%—#20–171
16gemini-3.1-pro-previewGoogle DeepMindsept. 2026$2 / $1277.1%—#9–36
17claude-opus-4-7Anthropicsept. 2026$5 / $2575.8%—#1–18
18gpt-5.4-2026-03-05OpenAIsept. 2026$2.50 / $1574.0%—#12–27
19deepseek-v4.1-flashDeepSeekoct. 2026$0.15 / $0.6072.9%—#14–30
20claude-opus-4-8Anthropicsept. 2026$5 / $2572.1%—#15–33
21gemini-3.5-flashGoogle DeepMindsept. 2026$1.50 / $972.1%—#10–26
22claude-opus-4-6Anthropicsept. 2026$5 / $2569.2%—#2–25
23grok-4.6xAIsept. 2026$2 / $667.1%—#15–52
24glm-5.3-flashZ.ai (Zhipu AI)aug. 2026$0.05 / $0.2065.8%—#26–59
25grok-4-20xAI—$1.25 / $2.5065.1%—#44–259
26deepseek-v4-flash-0731DeepSeekiul. 2026$0.15 / $0.1561.4%—#45–244
27grok-4.7xAIoct. 2026$2 / $661.4%—#23–72
28deepseek-v4-pro-0813DeepSeekaug. 2026$0.66 / $1.9861.3%—#40–130
29claude-sonnet-4-6Anthropicsept. 2026$3 / $1560.4%—#39–78
30gemini-3.6-flashGoogle DeepMindsept. 2026$0.75 / $3.7560.4%—#16–95
31kimi-k3Moonshotiun. 2026$0.24 / $11.9460.4%—#10–40
32gpt-5.6-lunaOpenAIsept. 2026$0.20 / $1.2059.5%—#30–147
33gpt-6-lunaOpenAIoct. 2026$0.10 / $0.5059.3%—#31–85
34gpt-5.2-pro-2025-12-11OpenAIdec. 2025$21 / $16854.2%—#172–262
35gpt-5.2-2025-12-11OpenAIdec. 2025$1.75 / $1452.9%—#34–93

Se afișează ARC-AGI-2 exact așa cum a publicat-o sursa — propriul scor și interval de încredere, fără mix. Selectează o a doua axă ca să construiești un scor compus ponderat.

Puținele lucruri din AI și cuantic care au contat cu adevărat săptămâna asta — rezumate cinstit.

Un e-mail pe săptămână. Dezabonare dintr-un clic. Fără urmărire, fără partajare.