Preseason
MatchesRankingsPrompts
GitHub
Preseason
MatchesRankingsPromptsMethodologyContact

© 2026 Preseason. All rights reserved.

PrivacyTerms
@betocmn
LLM Evals
Methodology

Promptfoo vs Arize AI

PromptfooPRPromptfoovsARArize AI
PromptfooArize AI
45%
55%

Leading: Arize AI (54.6%)

Statistics

MetricValue
Promptfoo wins123
Arize AI wins148
Abstains (no tool)237
Other tool chosen4020
Decisive cases271
Promptfoo win rate (unweighted)45.4%
95% CI39.6% - 51.3%
Promptfoo win rate (weighted)45.4%

Comments

Promptfoo

No comments yet

Verified critics can leave comments here.

Arize AI

No comments yet

Verified critics can leave comments here.

Model range breakdown

Model rangeTierPromptfooArize AINoneOtherA rate
DeepSeek Flash 3.2–4.1Mid512620483%
DeepSeek Pro 4–4 0813Frontier402102100%
DeepSeek R1Frontier04122240%
DevstralMid6161916027%
Gemini Flash 2.5–3.8Small14271218234%
Gemini Pro 2.5–3.1 PreviewFrontier5201719620%
GLM 5 Turbo–5.3Frontier25019196100%
GPT 5.4–6 AstraFrontier105234100%
GPT Coding 5.3 Codex–6.1 SolFrontier102237100%
GPT Mini 5.4 Mini–6 LunaMid1009218100%
HaikuSmall0242250%
Kimi 2.5–3Frontier161520494%
Llama MaverickFrontier015181980%
Llama ScoutSmall057291280%
MiMo Pro 2–2.6Frontier16017206100%
MiniMax M 2.7–3Frontier311222075%
Mistral SmallMid123520680%
Opus 4.6–5.5Frontier509225100%
Qwen CoderMid0142260%
Sonnet 4.6–5.5Frontier0011229n/a

Per-prompt breakdown

PromptTierPromptfooArize AINoneOtherA rate
ai-revenue-ops-copilotAdvanced729353519%
ai-support-agent-platformIntermediate1221654136%
ai-engineering-workflowAdvanced255113783%
ai-revenue-ops-copilotIntermediate1118554638%
ai-support-agent-platformBeginner5238047618%
ai-support-agent-platformAdvanced819555530%
ai-revenue-ops-copilotBeginner6201354523%
ai-engineering-workflowBeginner241826596%
ai-engineering-workflowIntermediate182115290%
ai-agent-applicationAdvanced28016320%
ai-agent-applicationIntermediate32116960%
ai-agent-applicationBeginner2040136100%