fedi-index: Agenten
Modelle als Agenten: Tool-Nutzung und mehrstufige Aufgaben – Arena-Agent-Sitzungen, τ²-bench- und Terminal-Bench-Läufe.
| # | Tier | Modell | fedi-index | Perzentil | Menschen bevorzugen | Löst Aufgaben | Preis-Leistung | Quellen | Aufschlüsselung des Werts |
|---|---|---|---|---|---|---|---|---|---|
| 1 | S | 1’530 ±9 | 100.0 | 1’559 | 1’469 | +40 | AT | › | |
| 2 | S | 1’506 ±7 | 99 | 1’522 | 1’468 | +73 | AT | › | |
| 3 | S | 1’502 ±8 | 98 | 1’520 | 1’461 | +65 | AT | › | |
| 4 | S | 1’500 ±3 | 96 | 1’513 | 1’466 | +108◆ | AT | › | |
| 5 | S | 1’494 ±9 | 95 | 1’507 | 1’465 | +49 | AT | › | |
| 6 | A | 1’492 ±8 | 94 | 1’506 | 1’463 | -24 | AT | › | |
| 7 | A | 1’491 ±6 | 93 | 1’504 | 1’460 | +9 | AT | › | |
| 8 | A | 1’487 ±7 | 91 | 1’500 | 1’455 | +42 | AT | › | |
| 9 | A | 1’486 ±6 | 90 | 1’500 | 1’453 | +92◆ | AT | › | |
| 10 | A | 1’482 ±4 | 89 | 1’491 | 1’460 | +60 | AT | › | |
| 11 | A | 1’481 ±8 | 88 | 1’489 | 1’462 | -28 | AT | › | |
| 12 | A | 1’477 ±3 | 86 | 1’487 | 1’452 | +98 | AT | › | |
| 13 | A | 1’476 ±4 | 85 | 1’487 | 1’448 | +47 | AT | › | |
| 14 | A | 1’468 ±5 | 84 | 1’468 | 1’469 | +44 | AτT | › | |
| 15 | A | 1’461 ±5 | 83 | 1’466 | 1’454 | +25 | AτT | › | |
| 16 | A | 1’461 ±5 | 82 | 1’469 | 1’443 | +5 | AT | › | |
| 17 | A | 1’461 ±4 | 80 | 1’476 | — | — | A | › | |
| 18 | B | 1’458 ±4 | 79 | 1’458 | 1’457 | +52 | AT | › | |
| 19 | B | 1’457 ±6 | 78 | 1’473 | — | — | A | › | |
| 20 | B | 1’456 ±4 | 76 | 1’457 | 1’454 | +36 | AτT | › | |
| 21 | B | 1’455 ±4 | 75 | 1’468 | 1’422 | — | AT | › | |
| 22 | B | 1’441 ±4 | 74 | 1’443 | 1’437 | +52 | AT | › | |
| 23 | B | 1’440 ±5 | 73 | 1’446 | 1’424 | +15 | AT | › | |
| 24 | B | 1’438 ±2 | 72 | 1’444 | 1’423 | +102 | AT | › | |
| 25 | B | 1’437 ±4 | 70 | 1’440 | 1’433 | +39 | AτT | › | |
| 26 | B | 1’437 ±5 | 69 | 1’439 | 1’432 | +49 | Aτ | › | |
| 27 | B | 1’437 ±3 | 68 | 1’440 | 1’428 | +65 | AT | › | |
| 28 | B | 1’435 ±3 | 67 | 1’445 | 1’410 | — | AT | › | |
| 29 | B | 1’435 ±3 | 65 | 1’447 | 1’404 | +101◆ | AT | › | |
| 30 | B | 1’434 ±4 | 64 | 1’438 | 1’426 | +104◆ | AT | › | |
| 31 | B | 1’433 ±4 | 63 | 1’447 | — | +30 | A | › | |
| 32 | B | 1’431 ±3 | 62 | 1’447 | 1’390 | +94◆ | AT | › | |
| 33 | B | 1’428 ±3 | 60 | 1’445 | 1’385 | +79 | AT | › | |
| 34 | B | 1’427 ±3 | 59 | 1’436 | 1’404 | +56 | AT | › | |
| 35 | B | 1’427 ±2 | 58 | 1’444 | 1’386 | +95 | AT | › | |
| 36 | B | 1’426 ±4 | 57 | 1’445 | 1’380 | +30 | AT | › | |
| 37 | B | 1’425 ±3 | 56 | 1’438 | 1’392 | — | AT | › | |
| 38 | B | 1’425 ±4 | 54 | 1’437 | 1’393 | +35 | AT | › | |
| 39 | B | 1’422 ±3 | 53 | 1’423 | 1’421 | +124◆ | AT | › | |
| 40 | B | 1’419 ±4 | 52 | 1’432 | 1’387 | +105◆ | AT | › | |
| 41 | B | 1’419 ±3 | 51 | 1’428 | 1’396 | +38 | AT | › | |
| 42 | C | 1’417 ±7 | 49 | — | 1’435 | +21 | τT | › | |
| 43 | C | 1’414 ±3 | 48 | 1’427 | 1’381 | +87 | AT | › | |
| 44 | C | 1’413 ±7 | 47 | — | 1’431 | +34 | τT | › | |
| 45 | C | 1’411 ±5 | 46 | 1’414 | 1’407 | +41 | AτT | › | |
| 46 | C | 1’411 ±6 | 44 | 1’423 | — | — | A | › | |
| 47 | C | 1’410 ±4 | 43 | 1’400 | 1’437 | +58 | Aτ | › | |
| 48 | C | 1’409 ±12 | 42 | — | 1’442 | +22 | τ | › | |
| 49 | C | 1’406 ±7 | 41 | 1’407 | 1’405 | +48 | AT | › | |
| 50 | C | 1’406 ±3 | 40 | 1’418 | — | — | A | › | |
| 51 | C | 1’404 ±6 | 38 | 1’417 | — | +35 | A | › | |
| 52 | C | 1’404 ±12 | 37 | — | 1’436 | +27 | τ | › | |
| 53 | C | 1’404 ±4 | 36 | 1’416 | — | +64 | A | › | |
| 54 | C | 1’403 ±12 | 35 | — | 1’435 | +35 | τ | › | |
| 55 | C | 1’396 ±11 | 33 | — | 1’426 | +1 | τ | › | |
| 56 | C | 1’391 ±6 | 32 | 1’403 | — | +46 | A | › | |
| 57 | C | 1’391 ±11 | 30 | — | 1’420 | +32 | τ | › | |
| 58 | C | 1’391 ±11 | 30 | — | 1’420 | +62 | τ | › | |
| 59 | C | 1’390 ±6 | 28 | 1’401 | — | +80 | A | › | |
| 60 | C | 1’389 ±11 | 27 | — | 1’417 | -8 | τ | › | |
| 61 | C | 1’388 ±4 | 26 | 1’399 | — | +63 | A | › | |
| 62 | D | 1’388 ±11 | 25 | — | 1’416 | -4 | τ | › | |
| 63 | D | 1’388 ±7 | 24 | 1’400 | — | +59 | A | › | |
| 64 | D | 1’388 ±12 | 22 | 1’401 | — | +42 | A | › | |
| 65 | D | 1’386 ±4 | 21 | 1’391 | 1’373 | +58 | AT | › | |
| 66 | D | 1’382 ±10 | 20 | — | 1’408 | +28 | τ | › | |
| 67 | D | 1’378 ±9 | 18 | — | 1’403 | -1 | τ | › | |
| 68 | D | 1’374 ±8 | 16 | — | 1’398 | +26 | τ | › | |
| 69 | D | 1’374 ±8 | 16 | — | 1’398 | +33 | τ | › | |
| 70 | D | 1’374 ±5 | 15 | 1’385 | — | — | A | › | |
| 71 | D | 1’370 ±5 | 14 | — | 1’396 | — | T | › | |
| 72 | D | 1’365 ±4 | 12 | — | 1’389 | +91 | T | › | |
| 73 | D | 1’364 ±8 | 11 | 1’375 | — | -20 | A | › | |
| 74 | D | 1’356 ±4 | 10 | — | 1’380 | — | T | › | |
| 75 | D | 1’353 ±4 | 9 | 1’362 | — | +48 | A | › | |
| 76 | D | 1’351 ±14 | 7 | 1’362 | — | +6 | A | › | |
| 77 | D | 1’350 ±4 | 6 | — | 1’372 | +43 | T | › | |
| 78 | D | 1’348 ±7 | 5 | 1’357 | — | +28 | A | › | |
| 79 | D | 1’347 ±9 | 4 | 1’356 | — | +50 | A | › | |
| 80 | D | 1’343 ±2 | 2 | — | 1’364 | +16 | T | › | |
| 81 | D | 1’341 ±1 | 1 | — | 1’362 | — | T | › | |
| 82 | D | 1’324 ±9 | 0 | 1’332 | — | — | A | › |
AArena (menschliche Stimmen)TTerminal-Benchττ²-bench (Agenten) ◆ Pareto-Front: nichts ist zugleich besser und günstiger
Der fedi-index ist ein Elo-Äquivalent auf fester Skala: innerhalb einer Methodikversion zwischen Monaten vergleichbar. Tiers S–D: nach Perzentil im Bereich und nicht besser, als das Konfidenzintervall zulässt (Grenzwerte in der Methodik). „Vorläufig“ – nur eine Quellenfamilie hat das Modell hier gemessen. Methodik
Auf deiner Website einbetten
Die Tabelle als Widget für deine Website: aktualisiert sich selbst und behält die Quellenangaben.
Kopiert
JSON
fedi-index © fedi.software, CC BY 4.0; Komponenten behalten die Lizenzen ihrer Quellen.