Preseason
MatchesRankingsPrompts
GitHub
Preseason
MatchesRankingsPromptsMethodologyContact

© 2026 Preseason. All rights reserved.

PrivacyTerms
@betocmn
AI / LLM Integration
Methodology

Anthropic vs Claude Code

AnthropicANAnthropicvsClaude CodeCLClaude Code
AnthropicClaude Code
46%
54%

Leading: Claude Code (54.0%)

Statistics

MetricValue
Anthropic wins167
Claude Code wins196
Abstains (no tool)128
Other tool chosen3497
Decisive cases363
Anthropic win rate (unweighted)46.0%
95% CI40.9% - 51.1%
Anthropic win rate (weighted)46.0%

Comments

Anthropic

No comments yet

Verified critics can leave comments here.

Claude Code

No comments yet

Verified critics can leave comments here.

Model range breakdown

Model rangeTierAnthropicClaude CodeNoneOtherA rate
DeepSeek Flash 3.2–4.1Mid11891805%
DeepSeek Pro 4–4 0813Frontier039690%
DeepSeek R1Frontier0023187n/a
DevstralMid0016169n/a
Gemini Flash 2.5–3.8Small62120275%
Gemini Pro 2.5–3.1 PreviewFrontier102209100%
GLM 5 Turbo–5.3Frontier314319118%
GPT 5.4–6 AstraFrontier54019856%
GPT Coding 5.3 Codex–6.1 SolFrontier32020760%
GPT Mini 5.4 Mini–6 LunaMid15120417%
HaikuSmall924008070%
Kimi 2.5–3Frontier118317558%
Llama MaverickFrontier0027179n/a
Llama ScoutSmall0011192n/a
MiMo Pro 2–2.6Frontier211419315%
MiniMax M 2.7–3Frontier2769010728%
Mistral SmallMid11018145100%
Opus 4.6–5.5Frontier01601960%
Qwen CoderMid201207100%
Sonnet 4.6–5.5Frontier24020733%

Per-prompt breakdown

PromptTierAnthropicClaude CodeNoneOtherA rate
ai-revenue-ops-copilotBeginner2245148833%
ai-revenue-ops-copilotIntermediate3420153363%
ai-revenue-ops-copilotAdvanced543452110%
ai-support-agent-platformIntermediate3110055076%
ai-agent-applicationAdvanced19192611150%
ai-agent-applicationIntermediate1523448939%
ai-support-agent-platformAdvanced1422155239%
ai-agent-applicationBeginner1595010463%
ai-support-agent-platformBeginner125154971%