b.buildnscale
ToolsAI benchmarksBuild logRoadmap
Open workspace
buildnscale

A little less friction.
A little more shipped.

Built in the open

Your toolkit

Daily Focus PlannerPomodoro + Task LogMarkdown NotesJSON ToolkitDowntime DetectorFreelancer Rate CalculatorHabit Streak TrackerWeekly Review Dashboard

Around here

AI benchmarksCompare AI modelsBenchmark methodologyBuild logRoadmapChangelogWrite for usBack up your workspacePrivacyTermsSuggest a tool ↗
© 2026 BuildnScale · Muhammad YousafFree to use. Built for your next small win.
BUILDNSCALE / AI BENCHMARKSEvidence for your next build
OverviewCodingAgentsPrice & performanceModelsCompareMethodologyUpdatesArchive

Agents, measured.

Inspect results for a specific evaluation. Source coverage, test suite, and prompting configuration matter as much as the score.

Arena published dataset · source published (oldest included board) 22 Oct 2026

Latest published overall boards for text, web development, and agents. Each rating has its own scale and publication date. Results are imported, not evaluated by BuildnScale. Update check overdue: no successful source check in the last eight hours. Displaying the last validated snapshot.

Retrieved 26 Oct 2026 · Last successful check 2026-10-26 22:34:24 UTC · View exact source ↗ · CC BY 4.0

Source: Arena / lmarena-ai leaderboard dataset, CC BY 4.0. Filtered to the latest overall boards and normalized by BuildnScale; evaluations were not rerun. No endorsement implied.

Loading explorer…

Published evidence. Clear sources. No sponsored rank placement.

How to read these results →

50 models · Arena Agent · Higher is better

Arena agent IPS score (overall). Interpret within the source methodology; not interchangeable with text or web development ratings. Ranks apply only to this source snapshot and selected metric. Arena ranks are preserved from the source; other ranks use exact scores. Small differences do not establish statistical superiority.

Arena Agent results from arena. Select models to compare.
CompareRankModel / evaluated labelArena AgentSource
#1Claude Opus 5.5 (Max)anthropic0.2077

95% CI: 0.1839 to 0.2315

1,751,647 observations · 12,057 sessions

Published 2026-10-23

Evidence
#2Claude Sonnet 5.5 (Max)anthropic0.1399

95% CI: 0.1095 to 0.1702

1,054,171 observations · 8,097 sessions

Published 2026-10-23

Evidence
#3GPT-6 Astra (Max)openai0.1396

95% CI: 0.1188 to 0.1604

1,132,157 observations · 14,567 sessions

Published 2026-10-23

Evidence
#4Claude Fable 5.1 (Max)anthropic0.1394

95% CI: 0.1201 to 0.1588

2,055,139 observations · 17,062 sessions

Published 2026-10-23

Evidence
#5GPT-6.1 Sol (Max)openai0.1364

95% CI: 0.1201 to 0.1527

1,343,948 observations · 16,092 sessions

Published 2026-10-23

Evidence
#6Gemini 4 Argon (High)google0.1175

95% CI: 0.1084 to 0.1266

4,974,530 observations · 45,840 sessions

Published 2026-10-23

Evidence
#7GPT-6 Sol (Max)openai0.0857

95% CI: 0.0685 to 0.1028

2,137,040 observations · 10,305 sessions

Published 2026-10-23

Evidence
#8Claude Opus 5 (Max)anthropic0.0822

95% CI: 0.0659 to 0.0985

3,331,922 observations · 24,440 sessions

Published 2026-10-23

Evidence
#9Claude Fable 5 (High)anthropic0.0775

95% CI: 0.0635 to 0.0916

2,664,967 observations · 43,567 sessions

Published 2026-10-23

Evidence
#10GPT-5.6 Sol (xHigh)openai0.0606

95% CI: 0.0472 to 0.0739

4,679,840 observations · 37,542 sessions

Published 2026-10-23

Evidence
#11Claude Sonnet 5 (High)anthropic0.0375

95% CI: 0.0205 to 0.0545

3,759,502 observations · 33,991 sessions

Published 2026-10-23

Evidence
#12Muse Wonder (Max)meta0.0331

95% CI: 0.0173 to 0.0488

1,498,241 observations · 11,134 sessions

Published 2026-10-23

Evidence
#13MiMo-V2.6-Proxiaomi0.0330

95% CI: 0.0250 to 0.0411

4,281,990 observations · 43,983 sessions

Published 2026-10-23

Evidence
#14Gemini 3.8 Flash (High)google0.0317

95% CI: 0.0210 to 0.0425

5,854,973 observations · 39,719 sessions

Published 2026-10-23

Evidence
#15Muse Spark 1.3 (Max)meta0.0305

95% CI: 0.0222 to 0.0388

7,934,311 observations · 78,017 sessions

Published 2026-10-23

Evidence
#16Grok 4.7 (xHigh)xai0.0302

95% CI: 0.0145 to 0.0460

2,906,258 observations · 16,639 sessions

Published 2026-10-23

Evidence
#17GPT-6 Luna (Max)openai0.0279

95% CI: 0.0237 to 0.0322

30,165,039 observations · 194,437 sessions

Published 2026-10-23

Evidence
#18DeepSeek-V4.1-Flash-20260910 (Max)deepseek0.0225

95% CI: 0.0191 to 0.0260

41,366,427 observations · 259,479 sessions

Published 2026-10-23

Evidence
#19MiMo-V2.6-Flashxiaomi0.0210

95% CI: 0.0078 to 0.0343

2,527,967 observations · 22,895 sessions

Published 2026-10-23

Evidence
#20Hy4 previewtencent0.0202

95% CI: 0.0138 to 0.0265

9,861,457 observations · 93,764 sessions

Published 2026-10-23

Evidence
#21GLM-5.2 (Max)zai0.0169

95% CI: 0.0088 to 0.0251

7,332,857 observations · 90,051 sessions

Published 2026-10-23

Evidence
#22Kimi K3 (Max)moonshot0.0157

95% CI: 0.0082 to 0.0232

13,076,043 observations · 139,480 sessions

Published 2026-10-23

Evidence
#23GLM-5.3 (Max)zai0.0122

95% CI: 0.0038 to 0.0205

9,782,984 observations · 88,117 sessions

Published 2026-10-23

Evidence
#24Claude Haiku 5.5 (High)anthropic0.0114

95% CI: 0.0011 to 0.0217

3,217,098 observations · 27,887 sessions

Published 2026-10-23

Evidence
#25Qwen3.8-Maxalibaba0.0113

95% CI: 0.0052 to 0.0174

7,288,963 observations · 75,307 sessions

Published 2026-10-23

Evidence

#1

Claude Opus 5.5 (Max)
0.2077Arena Agent

95% CI: 0.1839 to 0.2315

1,751,647 observations · 12,057 sessions

Published 2026-10-23

anthropicEvidence

#2

Claude Sonnet 5.5 (Max)
0.1399Arena Agent

95% CI: 0.1095 to 0.1702

1,054,171 observations · 8,097 sessions

Published 2026-10-23

anthropicEvidence

#3

GPT-6 Astra (Max)
0.1396Arena Agent

95% CI: 0.1188 to 0.1604

1,132,157 observations · 14,567 sessions

Published 2026-10-23

openaiEvidence

#4

Claude Fable 5.1 (Max)
0.1394Arena Agent

95% CI: 0.1201 to 0.1588

2,055,139 observations · 17,062 sessions

Published 2026-10-23

anthropicEvidence

#5

GPT-6.1 Sol (Max)
0.1364Arena Agent

95% CI: 0.1201 to 0.1527

1,343,948 observations · 16,092 sessions

Published 2026-10-23

openaiEvidence

#6

Gemini 4 Argon (High)
0.1175Arena Agent

95% CI: 0.1084 to 0.1266

4,974,530 observations · 45,840 sessions

Published 2026-10-23

googleEvidence

#7

GPT-6 Sol (Max)
0.0857Arena Agent

95% CI: 0.0685 to 0.1028

2,137,040 observations · 10,305 sessions

Published 2026-10-23

openaiEvidence

#8

Claude Opus 5 (Max)
0.0822Arena Agent

95% CI: 0.0659 to 0.0985

3,331,922 observations · 24,440 sessions

Published 2026-10-23

anthropicEvidence

#9

Claude Fable 5 (High)
0.0775Arena Agent

95% CI: 0.0635 to 0.0916

2,664,967 observations · 43,567 sessions

Published 2026-10-23

anthropicEvidence

#10

GPT-5.6 Sol (xHigh)
0.0606Arena Agent

95% CI: 0.0472 to 0.0739

4,679,840 observations · 37,542 sessions

Published 2026-10-23

openaiEvidence

#11

Claude Sonnet 5 (High)
0.0375Arena Agent

95% CI: 0.0205 to 0.0545

3,759,502 observations · 33,991 sessions

Published 2026-10-23

anthropicEvidence

#12

Muse Wonder (Max)
0.0331Arena Agent

95% CI: 0.0173 to 0.0488

1,498,241 observations · 11,134 sessions

Published 2026-10-23

metaEvidence

#13

MiMo-V2.6-Pro
0.0330Arena Agent

95% CI: 0.0250 to 0.0411

4,281,990 observations · 43,983 sessions

Published 2026-10-23

xiaomiEvidence

#14

Gemini 3.8 Flash (High)
0.0317Arena Agent

95% CI: 0.0210 to 0.0425

5,854,973 observations · 39,719 sessions

Published 2026-10-23

googleEvidence

#15

Muse Spark 1.3 (Max)
0.0305Arena Agent

95% CI: 0.0222 to 0.0388

7,934,311 observations · 78,017 sessions

Published 2026-10-23

metaEvidence

#16

Grok 4.7 (xHigh)
0.0302Arena Agent

95% CI: 0.0145 to 0.0460

2,906,258 observations · 16,639 sessions

Published 2026-10-23

xaiEvidence

#17

GPT-6 Luna (Max)
0.0279Arena Agent

95% CI: 0.0237 to 0.0322

30,165,039 observations · 194,437 sessions

Published 2026-10-23

openaiEvidence

#18

DeepSeek-V4.1-Flash-20260910 (Max)
0.0225Arena Agent

95% CI: 0.0191 to 0.0260

41,366,427 observations · 259,479 sessions

Published 2026-10-23

deepseekEvidence

#19

MiMo-V2.6-Flash
0.0210Arena Agent

95% CI: 0.0078 to 0.0343

2,527,967 observations · 22,895 sessions

Published 2026-10-23

xiaomiEvidence

#20

Hy4 preview
0.0202Arena Agent

95% CI: 0.0138 to 0.0265

9,861,457 observations · 93,764 sessions

Published 2026-10-23

tencentEvidence

#21

GLM-5.2 (Max)
0.0169Arena Agent

95% CI: 0.0088 to 0.0251

7,332,857 observations · 90,051 sessions

Published 2026-10-23

zaiEvidence

#22

Kimi K3 (Max)
0.0157Arena Agent

95% CI: 0.0082 to 0.0232

13,076,043 observations · 139,480 sessions

Published 2026-10-23

moonshotEvidence

#23

GLM-5.3 (Max)
0.0122Arena Agent

95% CI: 0.0038 to 0.0205

9,782,984 observations · 88,117 sessions

Published 2026-10-23

zaiEvidence

#24

Claude Haiku 5.5 (High)
0.0114Arena Agent

95% CI: 0.0011 to 0.0217

3,217,098 observations · 27,887 sessions

Published 2026-10-23

anthropicEvidence

#25

Qwen3.8-Max
0.0113Arena Agent

95% CI: 0.0052 to 0.0174

7,288,963 observations · 75,307 sessions

Published 2026-10-23

alibabaEvidence
Page 1 of 2

Source: Arena / lmarena-ai leaderboard dataset, CC BY 4.0. Filtered to the latest overall boards and normalized by BuildnScale; evaluations were not rerun. No endorsement implied.