Preseason
MatchesRankingsPrompts
GitHub
Preseason
MatchesRankingsPromptsMethodologyContact

© 2026 Preseason. All rights reserved.

PrivacyTerms
@betocmn
CMS
Methodology

Starlight vs Sphinx

STStarlightvsSPSphinx
StarlightSphinx
42%
58%

Leading: Sphinx (58.3%)

Insufficient data
This matchup has 12 decisive cases (minimum 30 required for publication).

Statistics

MetricValue
Starlight wins5
Sphinx wins7
Abstains (no tool)197
Other tool chosen2965
Decisive cases12
Starlight win rate (unweighted)41.7%
95% CI19.3% - 68.0%
Starlight win rate (weighted)41.7%

Comments

Starlight

No comments yet

Verified critics can leave comments here.

Sphinx

No comments yet

Verified critics can leave comments here.

Per-model breakdown

ModelTierStarlightSphinxNoneOtherA rate
Llama 4 ScoutSmall0761500%
DeepSeek V3.2Mid306117100%
MiniMax M2.7Frontier104119100%
Mistral Small 4Mid100158100%
Claude Haiku 4.5Small000168n/a
Claude Opus 4.6Frontier000126n/a
Claude Opus 4.8Frontier00042n/a
Claude Sonnet 4.6Frontier000168n/a
DeepSeek R1 0528Frontier0011156n/a
DeepSeek V4 FlashMid00140n/a
DeepSeek V4 ProFrontier00636n/a
Devstral 2 2512Mid000156n/a
Gemini 2.5 FlashSmall000126n/a
Gemini 2.5 ProFrontier002165n/a
Gemini 3.5 FlashSmall00636n/a
GLM 5 TurboFrontier004184n/a
GLM 5.2Frontier00042n/a
GPT 5.3 CodexFrontier007157n/a
GPT 5.4Frontier000126n/a
GPT 5.4 MiniMid007154n/a
GPT 5.5Frontier00042n/a
Kimi K2.5Frontier008116n/a
Kimi K2.7 CodeFrontier00536n/a
Llama 4 MaverickFrontier0026129n/a
MiMo V2 ProFrontier004122n/a
MiMo V2.5 ProFrontier00140n/a
MiniMax M3Frontier00042n/a
Qwen3 Coder NextMid0056112n/a

Per-prompt breakdown

PromptTierStarlightSphinxNoneOtherA rate
documentation-siteAdvanced231350240%
documentation-siteIntermediate143048820%
documentation-siteBeginner20124404100%
blog-platform-cmsAdvanced0030498n/a
blog-platform-cmsIntermediate000536n/a
blog-platform-cmsBeginner000537n/a