Preseason
MatchesRankingsPrompts
GitHub
Preseason
MatchesRankingsPromptsMethodologyContact

© 2026 Preseason. All rights reserved.

PrivacyTerms
@betocmn
LLM Observability
Methodology

Arize Phoenix vs Datadog

Arize PhoenixARArize PhoenixvsDatadogDADatadog
Arize PhoenixDatadog
49%
51%

Leading: Datadog (51.5%)

Statistics

MetricValue
Arize Phoenix wins50
Datadog wins53
Abstains (no tool)159
Other tool chosen4327
Decisive cases103
Arize Phoenix win rate (unweighted)48.5%
95% CI39.1% - 58.1%
Arize Phoenix win rate (weighted)48.5%

Comments

Arize Phoenix

No comments yet

Verified critics can leave comments here.

Datadog

No comments yet

Verified critics can leave comments here.

Model range breakdown

Model rangeTierArize PhoenixDatadogNoneOtherA rate
DeepSeek Flash 3.2–4.1Mid22423150%
DeepSeek Pro 4–4 0813Frontier001107n/a
DeepSeek R1Frontier002238n/a
DevstralMid3029177100%
Gemini Flash 2.5–3.8Small14441912%
Gemini Pro 2.5–3.1 PreviewFrontier009231n/a
GLM 5 Turbo–5.3Frontier000240n/a
GPT 5.4–6 AstraFrontier005235n/a
GPT Coding 5.3 Codex–6.1 SolFrontier002238n/a
GPT Mini 5.4 Mini–6 LunaMid106233100%
HaikuSmall0432290%
Kimi 2.5–3Frontier004222n/a
Llama MaverickFrontier0017221n/a
Llama ScoutSmall01361850%
MiMo Pro 2–2.6Frontier0182310%
MiniMax M 2.7–3Frontier1505217100%
Mistral SmallMid81422289%
Opus 4.6–5.5Frontier009231n/a
Qwen CoderMid2002217100%
Sonnet 4.6–5.5Frontier009231n/a

Per-prompt breakdown

PromptTierArize PhoenixDatadogNoneOtherA rate
ai-support-agent-platformAdvanced918156433%
ai-revenue-ops-copilotAdvanced1016355638%
ai-revenue-ops-copilotBeginner5133454328%
ai-support-agent-platformIntermediate81158489%
ai-engineering-workflowAdvanced71016588%
ai-support-agent-platformBeginner341756543%
ai-revenue-ops-copilotIntermediate501572100%
ai-engineering-workflowIntermediate300174100%
ai-agent-applicationAdvanced000175n/a
ai-agent-applicationBeginner0027151n/a
ai-agent-applicationIntermediate001174n/a
ai-engineering-workflowBeginner0074104n/a