From d5347d47a17797ccd0baae835693c37e5a757318 Mon Sep 17 00:00:00 2001
From: Omair Saleh <omair@quikcue.com>
Date: Thu, 5 Mar 2026 01:01:24 +0800
Subject: [PATCH] Switch AI to gpt-4.1-nano + add OpenAI key
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Model: gpt-4.1-nano (~$0.10/1M input, $0.40/1M output)
Priority: OpenAI (nano) → Gemini Flash (fallback)
Both keys now in server docker-compose.yml
---
 .../src/app/api/automations/ai/route.ts       | 57 ++++++++++---------
 pledge-now-pay-later/src/lib/ai.ts            | 13 +++--
 2 files changed, 36 insertions(+), 34 deletions(-)

diff --git a/pledge-now-pay-later/src/app/api/automations/ai/route.ts b/pledge-now-pay-later/src/app/api/automations/ai/route.ts
index ae6bef4..85140d3 100644
--- a/pledge-now-pay-later/src/app/api/automations/ai/route.ts
+++ b/pledge-now-pay-later/src/app/api/automations/ai/route.ts
@@ -2,46 +2,47 @@ import { NextRequest, NextResponse } from "next/server"
 import prisma from "@/lib/prisma"
 import { getUser } from "@/lib/session"
 
-const GEMINI_KEY = process.env.GEMINI_API_KEY
 const OPENAI_KEY = process.env.OPENAI_API_KEY
-const HAS_AI = !!(GEMINI_KEY || OPENAI_KEY)
+const GEMINI_KEY = process.env.GEMINI_API_KEY
+const HAS_AI = !!(OPENAI_KEY || GEMINI_KEY)
+const OPENAI_MODEL = "gpt-4.1-nano"
 
 async function chat(messages: Array<{ role: string; content: string }>, maxTokens = 600): Promise<string> {
   if (!HAS_AI) return ""
 
-  // Prefer Gemini (free), fall back to OpenAI
-  if (GEMINI_KEY) {
+  // Prefer OpenAI (gpt-4.1-nano), fall back to Gemini
+  if (OPENAI_KEY) {
     try {
-      const systemMsg = messages.find(m => m.role === "system")?.content || ""
-      const contents = messages.filter(m => m.role !== "system").map(m => ({
-        role: m.role === "assistant" ? "model" : "user",
-        parts: [{ text: m.content }],
-      }))
-      const res = await fetch(
-        `https://generativelanguage.googleapis.com/v1beta/models/gemini-2.0-flash:generateContent?key=${GEMINI_KEY}`,
-        {
-          method: "POST",
-          headers: { "Content-Type": "application/json" },
-          body: JSON.stringify({
-            systemInstruction: systemMsg ? { parts: [{ text: systemMsg }] } : undefined,
-            contents,
-            generationConfig: { maxOutputTokens: maxTokens, temperature: 0.8 },
-          }),
-        }
-      )
+      const res = await fetch("https://api.openai.com/v1/chat/completions", {
+        method: "POST",
+        headers: { "Content-Type": "application/json", Authorization: `Bearer ${OPENAI_KEY}` },
+        body: JSON.stringify({ model: OPENAI_MODEL, messages, max_tokens: maxTokens, temperature: 0.8 }),
+      })
       const data = await res.json()
-      return data.candidates?.[0]?.content?.parts?.[0]?.text || ""
+      return data.choices?.[0]?.message?.content || ""
     } catch { return "" }
   }
 
   try {
-    const res = await fetch("https://api.openai.com/v1/chat/completions", {
-      method: "POST",
-      headers: { "Content-Type": "application/json", Authorization: `Bearer ${OPENAI_KEY}` },
-      body: JSON.stringify({ model: "gpt-4o-mini", messages, max_tokens: maxTokens, temperature: 0.8 }),
-    })
+    const systemMsg = messages.find(m => m.role === "system")?.content || ""
+    const contents = messages.filter(m => m.role !== "system").map(m => ({
+      role: m.role === "assistant" ? "model" : "user",
+      parts: [{ text: m.content }],
+    }))
+    const res = await fetch(
+      `https://generativelanguage.googleapis.com/v1beta/models/gemini-2.0-flash:generateContent?key=${GEMINI_KEY}`,
+      {
+        method: "POST",
+        headers: { "Content-Type": "application/json" },
+        body: JSON.stringify({
+          systemInstruction: systemMsg ? { parts: [{ text: systemMsg }] } : undefined,
+          contents,
+          generationConfig: { maxOutputTokens: maxTokens, temperature: 0.8 },
+        }),
+      }
+    )
     const data = await res.json()
-    return data.choices?.[0]?.message?.content || ""
+    return data.candidates?.[0]?.content?.parts?.[0]?.text || ""
   } catch { return "" }
 }
 
diff --git a/pledge-now-pay-later/src/lib/ai.ts b/pledge-now-pay-later/src/lib/ai.ts
index 0f7815c..0e76758 100644
--- a/pledge-now-pay-later/src/lib/ai.ts
+++ b/pledge-now-pay-later/src/lib/ai.ts
@@ -4,9 +4,10 @@
  * Falls back to smart heuristics when no API key is set
  */
 
-const GEMINI_KEY = process.env.GEMINI_API_KEY
 const OPENAI_KEY = process.env.OPENAI_API_KEY
-const HAS_AI = !!(GEMINI_KEY || OPENAI_KEY)
+const GEMINI_KEY = process.env.GEMINI_API_KEY
+const HAS_AI = !!(OPENAI_KEY || GEMINI_KEY)
+const OPENAI_MODEL = "gpt-4.1-nano" // ~$0.10/1M input, ~$0.40/1M output
 
 interface ChatMessage {
   role: "system" | "user" | "assistant"
@@ -16,9 +17,9 @@ interface ChatMessage {
 async function chat(messages: ChatMessage[], maxTokens = 300): Promise<string> {
   if (!HAS_AI) return ""
 
-  // Prefer Gemini (free), fall back to OpenAI
-  if (GEMINI_KEY) return chatGemini(messages, maxTokens)
-  return chatOpenAI(messages, maxTokens)
+  // Prefer OpenAI (gpt-4.1-nano), fall back to Gemini
+  if (OPENAI_KEY) return chatOpenAI(messages, maxTokens)
+  return chatGemini(messages, maxTokens)
 }
 
 async function chatGemini(messages: ChatMessage[], maxTokens: number): Promise<string> {
@@ -63,7 +64,7 @@ async function chatOpenAI(messages: ChatMessage[], maxTokens: number): Promise<s
         "Content-Type": "application/json",
         Authorization: `Bearer ${OPENAI_KEY}`,
       },
-      body: JSON.stringify({ model: "gpt-4o-mini", messages, max_tokens: maxTokens, temperature: 0.7 }),
+      body: JSON.stringify({ model: OPENAI_MODEL, messages, max_tokens: maxTokens, temperature: 0.7 }),
     })
     const data = await res.json()
     return data.choices?.[0]?.message?.content || ""