Compare capability scores, evaluation time, and reference cost across models and reasoning efforts.

Evaluated on private datasetsMethodology

Data snapshot · 11/12/2026

Scoring dimension
About the scores

Backend engineering 20% · Frontend interaction 25% · Knowledge & reasoning 35% · Visual aesthetics 10% · Maintainability 10%. By default, each model shows its highest-scoring tested effort for the selected dimension.

Evaluation time and reference cost correspond to the selected dimension. They reflect benchmark runs, not everyday single-request latency or billing.

Explore evaluation tasksEvaluation method
Filters
Result scope
Rank byBy score
Ranking method

Preference rankings select the tested effort suited to that method; raw scores stay unchanged.
These are benchmark completion times, not everyday chat response speeds.

55 models · 109 configsBest effort per model

Scroll horizontally to view all scores and costs →

RankModelEval. time
Speed tok/sActions
01
Claude Fable 5.2NEW
Backend engineering97Frontend interaction96Knowledge & reasoning100Visual aesthetics83Maintainability73
94.0097961008373351m 33s$81.98…
02
gpt-6.2-astraNEW
Backend engineering97Frontend interaction97Knowledge & reasoning95Visual aesthetics83Maintainability78
93.00979795837894m 34s$16.64Cost calculation basisEstimated from this evaluation’s token usage at GPT-6 Astra Standard rates: $10 input, $50 output and $1 cached input per million tokens. This is a reference cost for GPT-6.2 Astra, not its quoted price or your actual bill.View reference rates ↗45
03
claude-opus-5-5
Backend engineering81Frontend interaction97Knowledge & reasoning100Visual aesthetics73Maintainability70
89.758197100737066m 3s$13.41…
04
Gemini 4 CarbonNEW
Backend engineering90Frontend interaction95Knowledge & reasoning95Visual aesthetics78Maintainability68
89.60909595786871m 3sN/A…
05
gpt-6-astra
Backend engineering95Frontend interaction94Knowledge & reasoning95Visual aesthetics75Maintainability61
89.35959495756183m 56s$7.01…
06
gpt-6.2-solNEW
Backend engineering96Frontend interaction91Knowledge & reasoning95Visual aesthetics73Maintainability64
88.90969195736447m 41sN/A…
07
gpt-6.1-sol
Backend engineering95Frontend interaction96Knowledge & reasoning90Visual aesthetics66Maintainability70
88.10959690667062m 9s$1.69…
08
claude-fable-5-1
Backend engineering83Frontend interaction91Knowledge & reasoning95Visual aesthetics80Maintainability68
87.40839195806879m 7s$26.15…
09
claude-sonnet-5-5
Backend engineering84Frontend interaction86Knowledge & reasoning85Visual aesthetics73Maintainability66
81.95848685736648m 32s$2.80…
10
gpt-6.1-terraNEW
Backend engineering89Frontend interaction84Knowledge & reasoning85Visual aesthetics64Maintainability61
81.05898485646121m 30s$0.85…
11
claude-opus-5
Backend engineering66Frontend interaction82Knowledge & reasoning95Visual aesthetics73Maintainability61
80.35668295736192m 12s$7.99…
12
muse-wonder
Backend engineering78Frontend interaction80Knowledge & reasoning90Visual aesthetics73Maintainability59
80.30788090735940m 8s$6.78…
13
k3.1-proNEW
Backend engineering81Frontend interaction91Knowledge & reasoning80Visual aesthetics68Maintainability64
80.158191806864109m 33s$5.0084
14
Gemini 4 Argon
Backend engineering79Frontend interaction82Knowledge & reasoning85Visual aesthetics70Maintainability59
78.95798285705946m 48sN/A…
15
Qwen 4 Max PreviewNEW
Backend engineering80Frontend interaction78Knowledge & reasoning90Visual aesthetics59Maintainability59
78.80807890595945m 38sN/A84
16
gpt-6-sol
Backend engineering77Frontend interaction92Knowledge & reasoning75Visual aesthetics75Maintainability66
78.75779275756639m 22s$1.10…
17
grok-4.7
Backend engineering87Frontend interaction74Knowledge & reasoning85Visual aesthetics66Maintainability55
77.758774856655196m 18s$5.25…
18
claude-opus-4-8
Backend engineering56Frontend interaction85Knowledge & reasoning90Visual aesthetics68Maintainability55
76.255685906855102m 10s$9.18…
19
grok-4.6
Backend engineering75Frontend interaction76Knowledge & reasoning85Visual aesthetics64Maintainability59
76.057576856459144m 16s$2.51…
20
hy4NEW
Backend engineering90Frontend interaction77Knowledge & reasoning75Visual aesthetics59Maintainability64
75.809077755964215m 2sN/A…
21
QoDuckNEW
Backend engineering70Frontend interaction80Knowledge & reasoning85Visual aesthetics64Maintainability55
75.65708085645544m 58sN/A95
22
deepseek-v4.1-flash-1017
Backend engineering69Frontend interaction91Knowledge & reasoning75Visual aesthetics61Maintainability53
74.20699175615364m 50s$0.45…
23
gpt-5.6-sol
Backend engineering83Frontend interaction73Knowledge & reasoning75Visual aesthetics66Maintainability64
74.10837375666487m 21s$3.25…
24
Egg AlphaNEW
Backend engineering65Frontend interaction78Knowledge & reasoning85Visual aesthetics61Maintainability51
73.45657885615132m 52sN/A95
24
MiniMax-M3.1-Flash
Backend engineering70Frontend interaction75Knowledge & reasoning80Visual aesthetics66Maintainability61
73.45707580666187m 15s$0.21…
26
muse-spark-1.3
Backend engineering68Frontend interaction85Knowledge & reasoning75Visual aesthetics59Maintainability47
71.70688575594729m 10s$0.066…
27
k3
Backend engineering65Frontend interaction82Knowledge & reasoning70Visual aesthetics68Maintainability59
70.706582706859158m 17s$3.28…
28
glm-5.3
Backend engineering67Frontend interaction72Knowledge & reasoning70Visual aesthetics66Maintainability64
68.906772706664153m 15s$1.83…
29
mimo-v2.6-pro
Backend engineering66Frontend interaction81Knowledge & reasoning65Visual aesthetics64Maintainability59
68.506681656459336m 3s$0.66…
30
space-bunny
Backend engineering57Frontend interaction74Knowledge & reasoning75Visual aesthetics59Maintainability55
67.55577475595562m 35s$0.000…
31
gpt-5.6-terra
Backend engineering78Frontend interaction76Knowledge & reasoning55Visual aesthetics61Maintainability61
66.057876556161133m 2s$1.94…
32
MiniMax-M3.1-Flash-Preview
Backend engineering71Frontend interaction53Knowledge & reasoning75Visual aesthetics70Maintainability53
66.00715375705354m 7sN/A…
33
deepseek-v4.1-flash
Backend engineering57Frontend interaction78Knowledge & reasoning65Visual aesthetics64Maintainability55
65.55577865645535m 40s$0.41…
34
claude-haiku-5-5
Backend engineering65Frontend interaction76Knowledge & reasoning60Visual aesthetics59Maintainability64
65.30657660596427m 38s$0.21…
35
hy4-preview
Backend engineering69Frontend interaction67Knowledge & reasoning65Visual aesthetics59Maintainability57
64.906967655957218m 41s$1.63…
36
Qwen 3.8 Max
Backend engineering65Frontend interaction67Knowledge & reasoning65Visual aesthetics61Maintainability57
64.306567656157254m 50s$4.9859
37
deepseek-v4-pro
Backend engineering64Frontend interaction60Knowledge & reasoning65Visual aesthetics70Maintainability59
63.45646065705993m 15s$1.15…
38
step-5
Backend engineering56Frontend interaction61Knowledge & reasoning75Visual aesthetics47Maintainability55
62.90566175475598m 22s$1.24…
39
k2.8
Backend engineering55Frontend interaction65Knowledge & reasoning70Visual aesthetics57Maintainability53
62.75556570575370m 58s$0.66…
40
LongCat-2.5NEW
Backend engineering44Frontend interaction62Knowledge & reasoning75Visual aesthetics55Maintainability53
61.35446275555394m 48s$0.68…
41
gpt-6-luna
Backend engineering57Frontend interaction78Knowledge & reasoning50Visual aesthetics68Maintainability59
61.10577850685963m 52s$0.061…
42
gemini-3.8-flash
Backend engineering60Frontend interaction84Knowledge & reasoning40Visual aesthetics66Maintainability59
59.50608440665925m 21s$0.13…
43
doubao-seed-2-1-pro
Backend engineering64Frontend interaction61Knowledge & reasoning55Visual aesthetics68Maintainability49
59.006461556849104m 54s$1.38…
44
Qwen 3.8 Flash
Backend engineering59Frontend interaction70Knowledge & reasoning50Visual aesthetics55Maintainability66
58.905970505566221m 2s$0.43115
45
Ling-3.2-flashNEW
Backend engineering59Frontend interaction54Knowledge & reasoning60Visual aesthetics61Maintainability57
58.105954606157—N/A…
46
deepseek-v4-flash
Backend engineering53Frontend interaction67Knowledge & reasoning55Visual aesthetics61Maintainability53
58.00536755615398m 25s$0.42…
47
deepseek-v4-flash-vision-exp
Backend engineering57Frontend interaction39Knowledge & reasoning70Visual aesthetics61Maintainability55
57.25573970615588m 43s$0.32…
48
mimo-v2.6-flash
Backend engineering54Frontend interaction67Knowledge & reasoning45Visual aesthetics59Maintainability57
54.905467455957386m 14s$0.28…
49
glm-5.3-flash
Backend engineering52Frontend interaction71Knowledge & reasoning40Visual aesthetics70Maintainability53
54.455271407053102m 55s$0.081…
50
Qwen 3.8 27b
Backend engineering59Frontend interaction85Knowledge & reasoning20Visual aesthetics66Maintainability59
52.555985206659104m 31s$1.79…
51
crouton-alphaNEW
Backend engineering45Frontend interaction64Knowledge & reasoning45Visual aesthetics59Maintainability53
51.9545644559538m 1sN/A553
52
gpt-5.6-luna
Backend engineering54Frontend interaction78Knowledge & reasoning25Visual aesthetics66Maintainability59
51.55547825665975m 23s$0.16…
53
MiniMax-M3
Backend engineering43Frontend interaction59Knowledge & reasoning50Visual aesthetics59Maintainability44
51.15435950594456m 33s$0.62…
54
gemini-3.7-flash
Backend engineering59Frontend interaction67Knowledge & reasoning25Visual aesthetics70Maintainability59
50.20596725705924m 19s$0.13…
55
Nemotron-3-Ultra
Backend engineering42Frontend interaction34Knowledge & reasoning40Visual aesthetics61Maintainability37
40.704234406137109m 30s$0.000…
Each model uses its highest-scoring effort here. Tied scores share a rank.

Evaluation methodology

Backend engineering 20% · Frontend interaction 25% · Knowledge & reasoning 35% · Visual aesthetics 10% · Maintainability 10%.