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
22.310
Ultima sincronizare
26 min
Toate sursele se sincronizează normal
  • Arena · Textacum 1 oră
  • Arena · WebDevacum 1 oră
  • Arena · Visionacum 1 oră
  • Arena · Agentacum 1 oră
  • GPQA Diamondacum 26 de minute
  • SWE-benchacum 26 de minute
  • Terminal-Benchacum 26 de minute
  • ARC-AGI-2acum 26 de minute
Agent Arena ×155 din 55 modele
ModelOrganizație± CIOscilațieAdaugă la comparație
1Claude Opus 5.5 (Max)anthropicoct. 2026$4 / $200.1890.161–0.216#2–11
2GPT 6.1 Sol (Max)openaioct. 2026$2 / $100.1480.128–0.168#1–15
3GPT-6 Astra (Max)openaisept. 2026$10 / $500.1410.118–0.164#1–18
4Claude Fable 5.1 (Max)anthropicsept. 2026$10 / $500.1400.120–0.160#2–21
5Gemini 4 Argon (High)googleoct. 2026$2 / $100.1320.119–0.146#1–10
6Claude Sonnet 5.5 (Max)anthropicoct. 2026$2 / $100.1300.099–0.161#4–26
7GPT 6 Sol (Max)openaioct. 2026$2 / $100.1120.089–0.135#7–53
8Claude Opus 5 (Max)anthropicsept. 2026$5 / $250.0870.070–0.104#5–11
9Claude Fable 5 (High)anthropicsept. 2026$10 / $500.0820.068–0.096#4–20
10GPT 5.6 Sol (xHigh)openaisept. 2026$4 / $200.0740.060–0.088#6–25
11Claude Opus 4.8 (High)anthropicsept. 2026$5 / $250.0710.056–0.087#15–33
12Muse Wonder (Max)metaoct. 2026—0.0490.032–0.066#14–56
13Claude Sonnet 5 (High)anthropicsept. 2026$2 / $100.0460.029–0.063#24–58
14MiMo-V2.6-Proxiaomisept. 2026$0.43 / $0.870.0400.029–0.050#18–64
15Muse Spark 1.3 (Max)metasept. 2026$1.25 / $4.250.0380.030–0.047#13–33
16Gemini 3.8 Flash (High)googleoct. 2026$0.83 / $4.130.0370.026–0.048#2–29
17GPT 5.5 (xHigh)openaisept. 2026$5 / $300.0330.020–0.046#9–34
18GPT 6 Luna (Max)openaioct. 2026$0.10 / $0.500.0330.028–0.037#30–85
19Grok 4.7 (xHigh)xaioct. 2026$2 / $60.0320.016–0.048#21–72
20Deepseek V4.1 Flash (Max)deepseekoct. 2026$0.15 / $0.600.0300.026–0.033#23–31
21Hy4 previewtencentaug. 2026$0.75 / $2.250.0270.020–0.034#32–136
22MiMo-V2.6-Flashxiaomisept. 2026$0.14 / $0.280.0260.012–0.040#19–88
23GLM 5.2 (Max)zaiiun. 2026$0.62 / $2.420.0240.015–0.032#9–49
24GLM 5.3 (Max)zaiaug. 2026$0.08 / $10.0200.012–0.028#18–52
25Kimi K3 (Max)moonshotiun. 2026$0.25 / $11.940.0190.012–0.027#10–39
26Qwen3.8 Maxalibabaoct. 2026$2 / $60.0190.013–0.025#12–40
27GPT 5.4 (High)openaisept. 2026$2.50 / $150.0160.005–0.027#12–29
28Claude Haiku 5.5 (High)anthropicoct. 2026$0.10 / $0.500.0150.005–0.024#43–277
29Grok 4.6 (xHigh)xaisept. 2026$2 / $60.009-0.002–0.020#15–52
30GPT 5.6 Terra (xHigh)openaisept. 2026$2 / $120.004-0.009–0.018#13–59
31DeepSeek V4 Pro (High) (0813)deepseekaug. 2026$0.66 / $1.980.004-0.014–0.021#17–54
32Grok 4.5xaisept. 2026$2 / $60.001-0.011–0.012#29–70
33Gemini 3.7 Flash (High)googleoct. 2026$0.75 / $3.75-0.007-0.016–0.002#9–57
34GLM 5.3 Flashzaiaug. 2026$0.05 / $0.20-0.011-0.018–-0.004#26–60
35Step 5 Previewstepfunsept. 2026$1 / $2.70-0.014-0.030–0.002#37–159

Se afișează Agent Arena 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.