feat: add new OpenAI models with proper temperature parameter handling

- Add GPT 4.1 series and o1/o3/o4 models with temperature compatibility fixes

- Remove gpt-5/gpt-5-mini models due to organization verification restrictions

- Fix 400 errors for models that only support default temperature values

src/lib/providers/index.ts

@@ -120,7 +120,11 @@ export const getAvailableChatModelProviders = async () => {
             model: new ChatOpenAI({
               apiKey: customOpenAiApiKey,
               modelName: customOpenAiModelName,
-              temperature: 0.7,
+              ...((() => {
+                const temperatureRestrictedModels = ['gpt-5-nano', 'o1', 'o3-mini', 'o4-mini'];
+                const isTemperatureRestricted = temperatureRestrictedModels.some(restrictedModel => customOpenAiModelName.includes(restrictedModel));
+                return isTemperatureRestricted ? {} : { temperature: 0.7 };
+              })()),
               configuration: {
                 baseURL: customOpenAiApiUrl,
               },