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
22.308
Ultima sincronizare
8 min
Toate sursele se sincronizează normal
  • Arena · Textacum 12 ore
  • Arena · WebDevacum 12 ore
  • Arena · Visionacum 12 ore
  • Arena · Agentacum 12 ore
  • GPQA Diamondacum 8 minute
  • SWE-benchacum 8 minute
  • Terminal-Benchacum 8 minute
  • ARC-AGI-2acum 8 minute
Urmarea instrucțiunilor ×1396 din 396 modele
ModelOrganizație± CIOscilațieAdaugă la comparație
1Gemini 4 Argon (High)googleoct. 2026$2 / $101,5271,512–1,542#1–10
2Claude Opus 5.5 (High)anthropicoct. 2026$4 / $201,5181,506–1,530#2–11
3Claude Opus 4.6 (High)anthropicsept. 2026$5 / $251,5121,507–1,517#2–25
4Claude Fable 5 (High)anthropicsept. 2026$10 / $501,5081,502–1,514#4–20
5Claude Opus 4.7 (High)anthropicsept. 2026$5 / $251,5021,497–1,508#1–18
6Claude Fable 5.1 (Max)anthropicsept. 2026$10 / $501,4991,489–1,508#2–21
7Claude Opus 5 (High)anthropicsept. 2026$5 / $251,4961,490–1,501#5–11
8Muse Wonder (Max)metaoct. 2026—1,4921,474–1,510#14–56
9claude-opus-4-8-highanthropicsept. 2026$5 / $251,4911,486–1,496#15–33
10gpt-5.6-sol-xhighopenaisept. 2026$4 / $201,4911,484–1,497#6–25
11Gemini 3.8 Flash (High)googleoct. 2026$0.83 / $4.131,4901,483–1,496#2–29
12Kimi K3 (Max)moonshotiun. 2026$0.28 / $11.941,4881,482–1,495#10–39
13gpt-6.1-sol-maxopenaioct. 2026$2 / $101,4871,472–1,501#1–15
14Gemini 3.7 Flash (High)googleoct. 2026$0.75 / $3.751,4851,478–1,493#9–57
15claude-opus-4-5-20251101-high-32kanthropicnov. 2025—1,4841,477–1,490#36–94
16claude-sonnet-5.5-xhighanthropicoct. 2026$2 / $101,4831,471–1,496#4–26
17Muse Spark 1.3 (Max)metasept. 2026$1.25 / $4.251,4831,475–1,491#13–33
18Gemini 3.1 Pro Previewgooglesept. 2026$2 / $121,4801,476–1,484#9–36
19gpt-5.5-highopenaisept. 2026$5 / $301,4781,472–1,483#9–34
20deepseek-v4.1-flash-maxdeepseekoct. 2026$0.15 / $0.601,4761,467–1,485#23–31
21glm-5.3-maxzaiaug. 2026$0.08 / $1.401,4751,467–1,482#18–52
22claude-sonnet-4-6anthropicsept. 2026$3 / $151,4751,469–1,480#38–77
23claude-opus-4-5-20251101anthropicnov. 2025$5 / $251,4741,469–1,479#20–62
24qwen3.8-maxalibabaoct. 2026$2 / $61,4741,467–1,481#12–40
25gpt-6-astra-maxopenaisept. 2026$10 / $501,4731,463–1,483#1–18
26Muse Spark 1.1metasept. 2026$1.25 / $4.251,4731,467–1,479#21–115
27gemini-3.6-flash-highgooglesept. 2026$0.75 / $3.751,4721,466–1,479#16–95
28Gemini 3 Progoogleian. 2026—1,4721,466–1,479#23–90
29glm-5.3-flashzaiaug. 2026$0.05 / $0.201,4721,465–1,478#26–60
30mimo-v2.6-proxiaomisept. 2026$0.43 / $0.871,4711,455–1,487#18–64
31Muse Spark 1.2 (xHigh)metaoct. 2026$1.25 / $4.251,4711,460–1,482#13–112
32gpt-5.4-highopenaisept. 2026$2.50 / $151,4711,465–1,476#12–29
33glm-5.2-maxzaiiun. 2026$0.62 / $2.421,4701,464–1,475#9–49
34mimo-v2.5-proxiaomiapr. 2026$0.43 / $0.871,4681,463–1,473#40–129
35qwen3.5-max-previewalibabaoct. 2026—1,4671,459–1,474#33–77

Se afișează Urmarea instrucțiunilor 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.