From a1f45fa33ac1129e743ca4f525c818d281052b8d Mon Sep 17 00:00:00 2001
From: Ajax Davis <thomasalwyndavis@gmail.com>
Date: Mon, 19 Jan 2026 05:20:05 +1000
Subject: [PATCH] fix: switch scenario evaluator to gpt-4.1-mini

Changed default evaluator from claude-3-5-haiku-latest to gpt-4.1-mini
since OPENAI_API_KEY is configured in production but ANTHROPIC_API_KEY
is not.
---
 apps/web/src/lib/scenarios/evaluate.ts | 9 ++++++---
 apps/web/src/lib/scenarios/execute.ts  | 2 +-
 2 files changed, 7 insertions(+), 4 deletions(-)

diff --git a/apps/web/src/lib/scenarios/evaluate.ts b/apps/web/src/lib/scenarios/evaluate.ts
index a2f74bb..e86b77e 100644
--- a/apps/web/src/lib/scenarios/evaluate.ts
+++ b/apps/web/src/lib/scenarios/evaluate.ts
@@ -24,7 +24,8 @@ export type EvaluatorModelId =
   | 'claude-3-5-sonnet-latest'
   | 'claude-3-5-haiku-latest'
   | 'gpt-4o'
-  | 'gpt-4o-mini';
+  | 'gpt-4o-mini'
+  | 'gpt-4.1-mini';
 
 /**
  * Get the model instance for an evaluator model ID
@@ -39,12 +40,14 @@ function getEvaluatorModel(modelId: EvaluatorModelId) {
       return openai('gpt-4o');
     case 'gpt-4o-mini':
       return openai('gpt-4o-mini');
+    case 'gpt-4.1-mini':
+      return openai('gpt-4.1-mini');
     default:
-      return anthropic('claude-3-5-haiku-latest');
+      return openai('gpt-4.1-mini');
   }
 }
 
-const DEFAULT_EVALUATOR: EvaluatorModelId = 'claude-3-5-haiku-latest';
+const DEFAULT_EVALUATOR: EvaluatorModelId = 'gpt-4.1-mini';
 
 /**
  * Evaluate if a scenario execution was successful
diff --git a/apps/web/src/lib/scenarios/execute.ts b/apps/web/src/lib/scenarios/execute.ts
index a9524ee..d2894ea 100644
--- a/apps/web/src/lib/scenarios/execute.ts
+++ b/apps/web/src/lib/scenarios/execute.ts
@@ -14,7 +14,7 @@ import {
   runAssertions,
 } from './evaluate';
 
-const DEFAULT_EVALUATOR: EvaluatorModelId = 'claude-3-5-haiku-latest';
+const DEFAULT_EVALUATOR: EvaluatorModelId = 'gpt-4.1-mini';
 const MAX_RETRIES = 1;
 
 interface ExecutionOptions {