K Koda Intelligence
trophyLeaderboard
LIVE · 4 BOARDS · UPDATED AUG 02, 2026

Who is actually winning

4 boards. 97 models ranked across 15 labs. One consensus verdict, verified weekly from the original sources.

CHAMPIONS · Artificial Analysis
#1
INDEX
Claude Opus 5 (max)
61
Anthropic
#2
INDEX
Claude Fable 5 (with fallback)
60
Anthropic
#3
INDEX
GPT-5.6 Sol (max)
59
OpenAI
// 01

The Consensus Board

The cross-board verdict. For each board a model family earns a family percentile (1 minus its canonical position), and the composite is the mean percentile across every board it appears on. Shown for models ranked on two or more boards. koda-leaderboard-composite-v1 · as of AUG 02, 2026 01:58 UTC.

#ModelCompositeLMArenaLiveBenchSWE-Bench VerifiedArtificial Analysis
1
claude-fable-5
98.6
#1#1·#2
2
Claude Opus 5 (max)
92.2
#6#4·#1
3
claude-opus-4-7-thinking
81.9
#3#9··
4
GPT-5.6 Sol Max Effort
81.5
#14#2·#3
5
GPT-5.6 Terra (max)
81.0
·#7·#5
6
Kimi K3 (max)
79.5
#12#5·#4
7
claude-opus-4-6-thinking
73.0
#2#16#4·
8
Gemini 3.1 Pro Preview High
72.7
#11#8··
9
GPT-5.5 Thinking xHigh Effort
70.5
#16#3··
10
Grok 4.5 (high)
70.2
·#12·#6
// 02

The Instruments

Every board, read straight from source. Bars scale to each board’s top score, so length is proportional and the exact figure sits in ink at the end.

LMArenaGeneral
AnthropicMetaGoogleMoonshot AI
1
claude-fable-5
1509
2
claude-opus-4-6-thinking
1505
3
claude-opus-4-7-thinking
1502
4
claude-opus-4-6
1497
5
claude-opus-4-7▲1
1492
6
claude-opus-5-highNEW
1492
7
claude-opus-5-maxNEW
1490
8
muse-spark-1.1▼3
1490
9
muse-spark▼2
1488
10
gemini-3-pro▼1
1486
11
gemini-3.1-pro-preview▼3
1485
12
kimi-k3-maxNEW
1485
SOURCE: arena.ai · SCRAPED AUG 02, 2026
LiveBenchReasoning
AnthropicOpenAIMoonshot AIGooglexAI
1
Claude Fable 5 Max Effort▲1
83
2
GPT-5.6 Sol Max Effort▼1
81
3
GPT-5.5 Thinking xHigh Effort▲1
80.2
4
Claude 5 Opus Thinking Max EffortNEW
80.1
5
Kimi K3▲2
79.2
6
GPT-5.4 Thinking xHigh Effort▲2
78
7
GPT-5.6 Terra Max Effort▼2
77.9
8
Gemini 3.1 Pro Preview High▲1
77
9
Claude 4.7 Opus Thinking xHigh Effort▲1
76.5
10
Claude 4.8 Opus Thinking Max EffortNEW
76.2
11
Claude Sonnet 5 xHigh Effort▲2
76
12
Grok 4.5▼1
75.8
SOURCE: livebench.ai · SCRAPED AUG 02, 2026
SWE-Bench VerifiedCoding
AnthropicGoogleMiniMaxOpenAIZ.AIMoonshot AIDeepSeek
1
Claude 4.5 Opus (high reasoning)
76.8
2
Gemini 3 Flash (high reasoning)
75.8
3
MiniMax M2.5 (high reasoning)
75.8
4
Claude Opus 4.6
75.6
5
GPT-5-2 Codex
72.8
6
GLM-5 (high reasoning)
72.8
7
GPT-5-2 (high reasoning)
72.8
8
Claude 4.5 Sonnet (high reasoning)▲1
71.4
9
Kimi K2.5 (high reasoning)▲1
70.8
10
DeepSeek V3.2 (high reasoning)▲1
70
11
Gemini 3 Pro▲1
69.6
12
Claude 4.5 Haiku (high reasoning)▲1
66.6
SOURCE: swebench.com · SCRAPED AUG 02, 2026
Artificial AnalysisIntelligence
AnthropicOpenAIMoonshot AIxAIZ.AIMetaGoogleDeepSeek
1
Claude Opus 5 (max)
61
2
Claude Fable 5 (with fallback)
60
3
GPT-5.6 Sol (max)
59
4
Kimi K3 (max)NEW
57
5
GPT-5.6 Terra (max)▲1
55
6
Grok 4.5 (high)▲1
54
7
Claude Sonnet 5 (max)▲1
53
8
GPT-5.6 Luna (max)▲1
51
9
GLM-5.2 (max)▲1
51
10
Muse Spark 1.1 (xhigh)▲1
51
11
Gemini 3.6 Flash▲1
50
12
DeepSeek V4 Flash 0731 (max)NEW
50
SOURCE: artificialanalysis.ai · SCRAPED AUG 02, 2026
// 03

Provider Dominance

Share of every top-20 canonical family slot across all 4 boards, by provider. One bar, 73 slots, the whole field at a glance.

OpenAI 25% (18)Anthropic 25% (18)Google 16% (12)Meta 5% (4)xAI 4% (3)DeepSeek 4% (3)Other 21% (15)
// 04

Head to Head

Pick any two canonical model families. We line them up on every board they share and show the exact configuration used. Boards where one is missing are skipped, not faked.

VS
// 05

Methodology

What each board measures

Composite koda-leaderboard-composite-v1 · as of AUG 02, 2026 01:58 UTC. Every figure is pulled from the source below and verified before it ships.

LMArenaarena.ai ↗SCRAPED AUG 02, 2026
Human preference rankings from blind A/B comparisons (7M+ votes)
LiveBenchlivebench.ai ↗SCRAPED AUG 02, 2026
Contamination-free benchmark with monthly-refreshed questions
SWE-Bench Verifiedswebench.com ↗SCRAPED AUG 02, 2026
Real-world software engineering task completion (Verified, 500 tasks)
Artificial Analysisartificialanalysis.ai ↗SCRAPED AUG 02, 2026
Artificial Analysis Intelligence Index -- aggregate of 9 evaluations

Different boards measure different things; the consensus view averages canonical-family percentiles, not scores. Each family contributes once per board using its best-ranked configuration.

Like what you see?

Get tomorrow's brief delivered to your inbox.

One email per day. Unsubscribe anytime.