pygpt-net 2.6.28__py3-none-any.whl → 2.6.30__py3-none-any.whl

pygpt_net/data/config/settings.json

@@ -133,6 +133,21 @@
         "advanced": false,
         "tab": "Google"
     },
+    "api_native_google": {
+        "section": "api_keys",
+        "type": "bool",
+        "slider": false,
+        "label": "settings.api_native_google",
+        "description": "settings.api_native_google.desc",
+        "value": true,
+        "min": null,
+        "max": null,
+        "multiplier": null,
+        "step": null,
+        "secret": false,
+        "advanced": false,
+        "tab": "Google"
+    },
     "api_key_anthropic": {
         "section": "api_keys",
         "type": "text",
@@ -683,20 +698,7 @@
         "multiplier": null,
         "step": null,
         "advanced": false
-    },
-    "ctx.attachment.verbose": {
-        "section": "files",
-        "type": "bool",
-        "slider": false,
-        "label": "settings.ctx.attachment.verbose",
-        "description": "settings.ctx.attachment.verbose.desc",
-        "value": false,
-        "min": null,
-        "max": null,
-        "multiplier": null,
-        "step": null,
-        "advanced": false
-    },
+    },    
     "ctx.attachment.img": {
         "section": "files",
         "type": "bool",
@@ -1276,7 +1278,22 @@
             {"1024x1024": "[DALL-E 3] 1024x1024"},
             {"1024x1024": "[DALL-E 2] 1024x1024"},
             {"512x512": "[DALL-E 2] 512x512"},
-            {"256x256": "[DALL-E 2] 256x256"}
+            {"256x256": "[DALL-E 2] 256x256"},
+            {"1024x1024": "[Imagen 3.0] 1024x1024"},
+            {"896x1280": "[Imagen 3.0] 896x1280"},
+            {"1280x896": "[Imagen 3.0] 1280x896"},
+            {"768x1408": "[Imagen 3.0] 768x1408"},
+            {"1408x768": "[Imagen 3.0] 1408x768"},
+            {"1024x1024": "[Imagen 4.0] 1024x1024"},
+            {"896x1280": "[Imagen 4.0] 896x1280"},
+            {"1280x896": "[Imagen 4.0] 1280x896"},
+            {"768x1408": "[Imagen 4.0] 768x1408"},
+            {"1408x768": "[Imagen 4.0] 1408x768"},
+            {"2048x2048": "[Imagen 4.0] 2048x2048"},
+            {"1792x2560": "[Imagen 4.0] 1792x2560"},
+            {"2560x1792": "[Imagen 4.0] 2560x1792"},
+            {"1536x2816": "[Imagen 4.0] 1536x2816"},
+            {"2816x1536": "[Imagen 4.0] 2816x1536"}
         ]
     },
     "img_quality": {
@@ -1610,6 +1627,34 @@
             "OpenAI Docs": "https://platform.openai.com/docs/guides/tools-file-search"
         }
     },
+    "remote_tools.google.web_search": {
+        "section": "remote_tools",
+        "type": "bool",
+        "slider": false,
+        "label": "settings.remote_tools.google.web_search",
+        "description": "settings.remote_tools.google.web_search.desc",
+        "value": true,
+        "min": null,
+        "max": null,
+        "multiplier": null,
+        "step": null,
+        "advanced": false,
+        "tab": "Google"
+    },
+    "remote_tools.google.code_interpreter": {
+        "section": "remote_tools",
+        "type": "bool",
+        "slider": false,
+        "label": "settings.remote_tools.google.code_interpreter",
+        "description": "settings.remote_tools.google.code_interpreter.desc",
+        "value": true,
+        "min": null,
+        "max": null,
+        "multiplier": null,
+        "step": null,
+        "advanced": false,
+        "tab": "Google"
+    },
     "llama.idx.list": {
         "section": "llama-index",
         "type": "dict",
@@ -2047,20 +2092,7 @@
         "step": 1,
         "advanced": false,
         "tab": "llama"
-    },
-    "agent.llama.verbose": {
-        "section": "agent",
-        "type": "bool",
-        "slider": false,
-        "label": "settings.agent.llama.verbose",
-        "value": false,
-        "min": 0,
-        "max": 0,
-        "multiplier": 1,
-        "step": 1,
-        "advanced": false,
-        "tab": "general"
-    },
+    },    
     "agent.idx.auto_retrieve": {
         "section": "agent",
         "type": "bool",
@@ -2259,10 +2291,10 @@
         "advanced": false
     },
     "debug": {
-        "section": "developer",
+        "section": "debug",
         "type": "bool",
         "slider": false,
-        "label": "settings.developer.debug",
+        "label": "settings.debug.show_menu",
         "value": false,
         "min": 0,
         "max": 0,
@@ -2270,8 +2302,27 @@
         "step": 1,
         "advanced": false
     },
+    "log.level": {
+        "section": "debug",
+        "description": "Tip: Running application with --debug=1 or --debug=2 command line arguments overwrites this settings and force enables logging to %workdir%/app.log file. Log levels: 1 = INFO, 2 = DEBUG",
+        "type": "combo",
+        "slider": false,
+        "label": "Log Level",
+        "value": "",
+        "min": null,
+        "max": null,
+        "multiplier": null,
+        "step": null,
+        "advanced": false,
+        "keys": [
+            {"error": "ERROR (default)"},
+            {"warning": "WARNING"},
+            {"info": "INFO"},
+            {"debug": "DEBUG"}
+        ]
+    },
     "log.ctx": {
-        "section": "developer",
+        "section": "debug",
         "type": "bool",
         "slider": false,
         "label": "Log and debug context",
@@ -2283,7 +2334,7 @@
         "advanced": false
     },
     "log.events": {
-        "section": "developer",
+        "section": "debug",
         "type": "bool",
         "slider": false,
         "label": "Log and debug events",
@@ -2295,7 +2346,7 @@
         "advanced": false
     },
     "log.plugins": {
-        "section": "developer",
+        "section": "debug",
         "type": "bool",
         "slider": false,
         "label": "Log plugin usage to console",
@@ -2307,7 +2358,7 @@
         "advanced": false
     },
     "log.dalle": {
-        "section": "developer",
+        "section": "debug",
         "type": "bool",
         "slider": false,
         "label": "Log DALL-E usage to console",
@@ -2318,11 +2369,11 @@
         "step": null,
         "advanced": false
     },
-    "log.llama": {
-        "section": "developer",
+    "ctx.attachment.verbose": {
+        "section": "debug",
         "type": "bool",
         "slider": false,
-        "label": "Log LlamaIndex usage to console",
+        "label": "Log attachments usage to console",
         "value": false,
         "min": null,
         "max": null,
@@ -2330,11 +2381,24 @@
         "step": null,
         "advanced": false
     },
-    "log.assistants": {
-        "section": "developer",
+    "agent.llama.verbose": {
+        "section": "debug",
         "type": "bool",
         "slider": false,
-        "label": "Log Assistants usage to console",
+        "label": "Log Agents usage to console",
+        "value": false,
+        "min": 0,
+        "max": 0,
+        "multiplier": 1,
+        "step": 1,
+        "advanced": false,
+        "tab": "general"
+    },
+    "log.llama": {
+        "section": "debug",
+        "type": "bool",
+        "slider": false,
+        "label": "Log LlamaIndex usage to console",
         "value": false,
         "min": null,
         "max": null,
@@ -2342,25 +2406,18 @@
         "step": null,
         "advanced": false
     },
-    "log.level": {
-        "section": "developer",
-        "description": "Tip: Running application with --debug=1 or --debug=2 command line arguments overwrites this settings and force enables logging to %workdir%/app.log file. Log levels: 1 = INFO, 2 = DEBUG",
-        "type": "combo",
+    "log.assistants": {
+        "section": "debug",
+        "type": "bool",
         "slider": false,
-        "label": "Log Level",
-        "value": "",
+        "label": "Log Assistants usage to console",
+        "value": false,
         "min": null,
         "max": null,
         "multiplier": null,
         "step": null,
-        "advanced": false,
-        "keys": [
-            {"error": "ERROR (default)"},
-            {"warning": "WARNING"},
-            {"info": "INFO"},
-            {"debug": "DEBUG"}
-        ]
-    },
+        "advanced": false
+    },    
     "access.voice_control": {
         "section": "access",
         "type": "bool",

pygpt_net/data/config/settings_section.json

@@ -47,7 +47,7 @@
     "updates": {
         "label": "settings.section.updates"
     },
-    "developer": {
-        "label": "settings.section.developer"
+    "debug": {
+        "label": "settings.section.debug"
     }
 }

pygpt_net/data/locale/locale.de.ini

@@ -1174,9 +1174,9 @@ settings.ctx.sources = Zeige Llama-Indexquellen
 settings.ctx.sources.desc = Falls aktiviert, werden die genutzten Quellen in der Antwort angezeigt (falls verfügbar, funktioniert nicht im gestreamten Chat)
 settings.ctx.use_extra = Verwenden Sie zusätzlichen Kontextoutput
 settings.ctx.use_extra.desc = Wenn aktiviert, wird die einfache Textausgabe (falls verfügbar) aus den Befehlsausgaben zusätzlich zur JSON-Ausgabe angezeigt.
+settings.debug.show_menu = Debug-Menü anzeigen
 settings.defaults.app.confirm = Werksseitige App-Einstellungen laden?
 settings.defaults.user.confirm = Aktuelle Änderungen rückgängig machen?
-settings.developer.debug = Debug-Menü anzeigen
 settings.dict.delete.confirm = Eintrag aus der Liste entfernen?
 settings.download.dir = Verzeichnis für Dateidownloads
 settings.download.dir.desc = Unterordner für heruntergeladene Dateien, z.B. im Assistentenmodus, innerhalb von "data"
@@ -1336,7 +1336,7 @@ settings.section.audio.cache = Cache
 settings.section.audio.device = Geräte
 settings.section.audio.options = Optionen
 settings.section.ctx = Kontext
-settings.section.developer = Entwickler
+settings.section.debug = Fehlerbehebung
 settings.section.files = Dateien und Anhänge
 settings.section.general = Allgemein
 settings.section.images = Bilder

pygpt_net/data/locale/locale.en.ini

@@ -1118,6 +1118,8 @@ settings.api_key.voyage = VoyageAI API KEY
 settings.api_key.voyage.desc = Required for the Voyage API - embeddings for Anthropic and DeepSeek API.
 settings.api_key.xai = xAI API KEY
 settings.api_key.xai.desc = Required for the xAI API and Grok models.
+settings.api_native_google = Use native API SDK
+settings.api_native_google.desc = Use native GenAI SDK instead of compatible OpenAI client
 settings.api_proxy = Proxy address
 settings.api_proxy.desc = Optional, proxy for OpenAI API, e.g. http://proxy.example.com or socks5://user:pass@host:port
 settings.api_use_responses = Use Responses API in Chat mode
@@ -1192,9 +1194,9 @@ settings.ctx.sources = Show LlamaIndex sources
 settings.ctx.sources.desc = If enabled, sources used will be displayed in the response (if available, it will not work in streamed chat)
 settings.ctx.use_extra = Use extra context output
 settings.ctx.use_extra.desc = If enabled, plain text output (if available) from command results will be displayed alongside the JSON output.
+settings.debug.show_menu = Show debug menu
 settings.defaults.app.confirm = Load factory app settings?
 settings.defaults.user.confirm = Undo current changes?
-settings.developer.debug = Show debug menu
 settings.dict.delete.confirm = Remove item from list?
 settings.download.dir = Directory for file downloads
 settings.download.dir.desc = Subdirectory for downloaded files, e.g. in Assistants mode, inside "data"
@@ -1324,6 +1326,10 @@ settings.remote_tools.file_search = File search
 settings.remote_tools.file_search.args = File search vector store IDs
 settings.remote_tools.file_search.args.desc = Vector store IDs, separated by comma (,)
 settings.remote_tools.file_search.desc = Enable `file_search` remote tool in Chat mode / via OpenAI Responses API.
+settings.remote_tools.google.code_interpreter = Code Interpreter
+settings.remote_tools.google.code_interpreter.desc = Enable Code Interpreter remote tool in Chat mode.
+settings.remote_tools.google.web_search = Google Web Search
+settings.remote_tools.google.web_search.desc = Enable Google Search remote tool in Chat mode.
 settings.remote_tools.image = Image generation
 settings.remote_tools.image.desc = Enable `image_generation` remote tool in Chat mode / via OpenAI Responses API.
 settings.remote_tools.mcp = Remote MCP
@@ -1360,7 +1366,7 @@ settings.section.audio.cache = Cache
 settings.section.audio.device = Devices
 settings.section.audio.options = Options
 settings.section.ctx = Context
-settings.section.developer = Developer
+settings.section.debug = Debug
 settings.section.files = Files and attachments
 settings.section.general = General
 settings.section.images = Images
@@ -1376,6 +1382,7 @@ settings.section.model = Models
 settings.section.personalize = Personalize
 settings.section.prompts = Prompts
 settings.section.remote_tools = Remote tools
+settings.section.remote_tools.google = Google
 settings.section.remote_tools.openai = OpenAI
 settings.section.tab.general = General
 settings.section.updates = Updates

pygpt_net/data/locale/locale.es.ini

@@ -1175,9 +1175,9 @@ settings.ctx.sources = Mostrar fuentes del índice Llama
 settings.ctx.sources.desc = Si está habilitado, las fuentes utilizadas se mostrarán en la respuesta (si están disponibles, no funcionará en el chat en vivo)
 settings.ctx.use_extra = Usar salida de contexto extra
 settings.ctx.use_extra.desc = Si está habilitado, la salida de texto sin formato (si está disponible) de los resultados de los comandos se mostrará junto con la salida JSON.
+settings.debug.show_menu = Mostrar menú de depuración
 settings.defaults.app.confirm = ¿Cargar ajustes predeterminados de la aplicación?
 settings.defaults.user.confirm = ¿Deshacer cambios actuales?
-settings.developer.debug = Mostrar menú de depuración
 settings.dict.delete.confirm = ¿Eliminar elemento de la lista?
 settings.download.dir = Directorio para descargas de archivos
 settings.download.dir.desc = Subdirectorio para archivos descargados, por ejemplo, en modo Asistentes, dentro de "data"
@@ -1337,7 +1337,7 @@ settings.section.audio.cache = Caché
 settings.section.audio.device = Dispositivos
 settings.section.audio.options = Opciones
 settings.section.ctx = Contexto
-settings.section.developer = Desarrollador
+settings.section.debug = Depuración
 settings.section.files = Archivos y adjuntos
 settings.section.general = General
 settings.section.images = Imágenes

pygpt_net/data/locale/locale.fr.ini

@@ -1174,9 +1174,9 @@ settings.ctx.sources = Afficher les sources de l'index Llama
 settings.ctx.sources.desc = Si activé, les sources utilisées seront affichées dans la réponse (si disponibles, cela ne fonctionnera pas dans le chat en continu)
 settings.ctx.use_extra = Utiliser une sortie de contexte supplémentaire
 settings.ctx.use_extra.desc = Si activé, la sortie en texte brut (si disponible) des résultats de commande sera affichée aux côtés de la sortie JSON.
+settings.debug.show_menu = Afficher le menu de débogage
 settings.defaults.app.confirm = Charger les réglages par défaut de l'application ?
 settings.defaults.user.confirm = Annuler les modifications actuelles ?
-settings.developer.debug = Afficher le menu de débogage
 settings.dict.delete.confirm = Supprimer l'élément de la liste ?
 settings.download.dir = Répertoire pour les téléchargements de fichiers
 settings.download.dir.desc = Sous-répertoire pour les fichiers téléchargés, par exemple en mode Assistants, à l'intérieur de "data"
@@ -1336,7 +1336,7 @@ settings.section.audio.cache = Cache
 settings.section.audio.device = Appareils
 settings.section.audio.options = Options
 settings.section.ctx = Contexte
-settings.section.developer = Développeur
+settings.section.debug = Débogage
 settings.section.files = Fichiers et pièces jointes
 settings.section.general = Général
 settings.section.images = Images

pygpt_net/data/locale/locale.it.ini

@@ -1174,9 +1174,9 @@ settings.ctx.sources = Mostra le fonti dell'indice Llama
 settings.ctx.sources.desc = Se abilitato, le fonti utilizzate saranno mostrate nella risposta (se disponibili, non funzionerà nella chat in streaming)
 settings.ctx.use_extra = Usa output di contesto extra
 settings.ctx.use_extra.desc = Se abilitato, l'output di testo normale (se disponibile) dai risultati dei comandi sarà visualizzato accanto all'output JSON.
+settings.debug.show_menu = Mostra menu debug
 settings.defaults.app.confirm = Caricare le impostazioni predefinite dell'applicazione?
 settings.defaults.user.confirm = Annullare le modifiche correnti?
-settings.developer.debug = Mostra menu debug
 settings.dict.delete.confirm = Rimuovere l'elemento dall'elenco?
 settings.download.dir = Directory per il download dei file
 settings.download.dir.desc = Sottodirectory per i file scaricati, ad esempio in modalità Assistenti, all'interno di "data"
@@ -1336,7 +1336,7 @@ settings.section.audio.cache = Cache
 settings.section.audio.device = Dispositivi
 settings.section.audio.options = Opzioni
 settings.section.ctx = Contesto
-settings.section.developer = Sviluppatore
+settings.section.debug = Debug
 settings.section.files = File e allegati
 settings.section.general = Generale
 settings.section.images = Immagini

pygpt_net/data/locale/locale.pl.ini

@@ -1118,7 +1118,7 @@ settings.audio.input.backend.desc = Wybierz backend dla wejścia audio.
 settings.audio.input.channels = Kanały
 settings.audio.input.channels.desc = Kanały wejściowe, domyślnie: 1
 settings.audio.input.continuous = Ciągłe Nagrywanie Dźwięku (Kawałki)
-settings.audio.input.continuous.desc = Włącz nagrywanie w kawałkach dla długich nagrań audio w notatniku (notatki głosowe).
+settings.audio.input.continuous.desc = Włącz nagrywanie w kawałkach dla długich nagrań audio w notatniku (notatki głosowe).
 settings.audio.input.device = Urządzenie do wejścia audio
 settings.audio.input.device.desc = Wybierz urządzenie do wejścia mikrofonu.
 settings.audio.input.rate = Częstotliwość próbkowania
@@ -1175,9 +1175,9 @@ settings.ctx.sources = Pokaż źródła LlamaIndex
 settings.ctx.sources.desc = Jeśli opcja jest włączona, wykorzystane źródła będą wyświetlane w odpowiedzi (jeśli dostępne, nie zadziała w czacie z wł. opcją stream)
 settings.ctx.use_extra = Używaj dodatkowego kontekstu outputu
 settings.ctx.use_extra.desc = Jeśli włączone, zwykły tekst outputu (jeśli dostępny) z wyników poleceń będzie wyświetlany obok outputu JSON.
+settings.debug.show_menu = Pokaż menu debugowania
 settings.defaults.app.confirm = Wczytać fabryczne ustawienia aplikacji?
 settings.defaults.user.confirm = Przywrócić dokonane zmiany?
-settings.developer.debug = Pokaż menu debugowania
 settings.dict.delete.confirm = Usunąć pozycję z listy?
 settings.download.dir = Katalog na pliki do pobrania
 settings.download.dir.desc = Podkatalog na pobrane pliki, np. w trybie Asystentów, wewnątrz "data"
@@ -1337,7 +1337,7 @@ settings.section.audio.cache = Pamięć podręczna
 settings.section.audio.device = Urządzenia
 settings.section.audio.options = Opcje
 settings.section.ctx = Kontekst
-settings.section.developer = Deweloper
+settings.section.debug = Debugowanie
 settings.section.files = Pliki i załączniki
 settings.section.general = Ogólne
 settings.section.images = Obrazy

pygpt_net/data/locale/locale.uk.ini

@@ -1174,9 +1174,9 @@ settings.ctx.sources = Показати джерела індексу Llama
 settings.ctx.sources.desc = Якщо включено, використані джерела будуть відображатися в відповіді (якщо доступно, не працюватиме в потоковому чаті)
 settings.ctx.use_extra = Використовувати додатковий контекст виводу
 settings.ctx.use_extra.desc = Якщо увімкнено, звичайний текстовий вивід (якщо доступний) з результатів команд буде відображений поруч з JSON виводом.
+settings.debug.show_menu = Показати меню налагодження
 settings.defaults.app.confirm = Завантажити заводські налаштування додатку?
 settings.defaults.user.confirm = Відмінити поточні зміни?
-settings.developer.debug = Показати меню налагодження
 settings.dict.delete.confirm = Видалити елемент зі списку?
 settings.download.dir = Директорія для завантаження файлів
 settings.download.dir.desc = Піддиректорія для завантажених файлів, наприклад, у режимі помічників, всередині "data"
@@ -1336,7 +1336,7 @@ settings.section.audio.cache = Кеш
 settings.section.audio.device = Пристрої
 settings.section.audio.options = Параметри
 settings.section.ctx = Контекст
-settings.section.developer = Розробник
+settings.section.debug = Налагодження
 settings.section.files = Файли та вкладення
 settings.section.general = Загальні
 settings.section.images = Зображення

pygpt_net/data/locale/locale.zh.ini

@@ -1174,9 +1174,9 @@ settings.ctx.sources = 显示Llama索引源
 settings.ctx.sources.desc = 如果启用，使用的源将在回应中显示（如果可用，不适用于流式聊天）
 settings.ctx.use_extra = 使用额外的上下文输出
 settings.ctx.use_extra.desc = 如果启用，将在命令结果的 JSON 输出旁边显示纯文本输出（如果有）。
+settings.debug.show_menu = 显示调试菜单
 settings.defaults.app.confirm = 加载出厂应用程序设置？
 settings.defaults.user.confirm = 撤销当前更改？
-settings.developer.debug = 显示调试菜单
 settings.dict.delete.confirm = 从列表中移除项目？
 settings.download.dir = 文件下载目录
 settings.download.dir.desc = 下载文件的子目录，例如在助手模式下，位于 "data" 内部
@@ -1336,7 +1336,7 @@ settings.section.audio.cache = 缓存
 settings.section.audio.device = 设备
 settings.section.audio.options = 选项
 settings.section.ctx = 上下文
-settings.section.developer = 開發者
+settings.section.debug = 调试
 settings.section.files = 文件和附件
 settings.section.general = 一般
 settings.section.images = 圖像

pygpt_net/item/model.py

@@ -6,7 +6,7 @@
 # GitHub:  https://github.com/szczyglis-dev/py-gpt   #
 # MIT License                                        #
 # Created By  : Marcin Szczygliński                  #
-# Updated Date: 2025.08.23 15:00:00                  #
+# Updated Date: 2025.08.28 09:00:00                  #
 # ================================================== #
 
 import json
@@ -253,7 +253,17 @@ class ModelItem:
 
         :return: True if supports image input
         """
-        if MODE_VISION in self.mode or MULTIMODAL_IMAGE in self.input:
+        if MULTIMODAL_IMAGE in self.input:
+            return True
+        return False
+
+    def is_image_output(self) -> bool:
+        """
+        Check if model supports image output
+
+        :return: True if supports image output
+        """
+        if "image" in self.output or MODE_VISION in self.mode:
             return True
         return False
 
@@ -263,7 +273,17 @@ class ModelItem:
 
         :return: True if supports audio input
         """
-        if MODE_AUDIO in self.mode or MULTIMODAL_AUDIO in self.input:
+        if MULTIMODAL_AUDIO in self.input:
+            return True
+        return False
+
+    def is_audio_output(self) -> bool:
+        """
+        Check if model supports audio output
+
+        :return: True if supports audio output
+        """
+        if MULTIMODAL_AUDIO in self.output:
             return True
         return False
 

pygpt_net/plugin/openai_dalle/plugin.py

@@ -6,7 +6,7 @@
 # GitHub:  https://github.com/szczyglis-dev/py-gpt   #
 # MIT License                                        #
 # Created By  : Marcin Szczygliński                  #
-# Updated Date: 2025.07.30 00:00:00                  #
+# Updated Date: 2025.08.28 09:00:00                  #
 # ================================================== #
 
 from pygpt_net.core.types import (
@@ -42,8 +42,8 @@ class Plugin(BasePlugin):
         ]
         self.allowed_modes = [
             MODE_CHAT,
-            MODE_LANGCHAIN,
-            MODE_VISION,
+            # MODE_LANGCHAIN,
+            # MODE_VISION,
             MODE_LLAMA_INDEX,
             MODE_ASSISTANT,
             MODE_AGENT,
@@ -166,7 +166,7 @@ class Plugin(BasePlugin):
                     sync = False
                     if self.window.core.config.get("mode") in [MODE_AGENT_LLAMA, MODE_AGENT_OPENAI]:
                         sync = True
-                    self.window.core.gpt.image.generate(bridge_context, extra, sync)  # force inline mode, async call
+                    self.window.core.api.openai.image.generate(bridge_context, extra, sync)  # force inline mode, async call
             except Exception as e:
                 self.log("Error: " + str(e))
                 return

pygpt_net/plugin/openai_vision/plugin.py

@@ -6,7 +6,7 @@
 # GitHub:  https://github.com/szczyglis-dev/py-gpt   #
 # MIT License                                        #
 # Created By  : Marcin Szczygliński                  #
-# Updated Date: 2025.08.15 23:00:00                  #
+# Updated Date: 2025.08.28 09:00:00                  #
 # ================================================== #
 
 from pygpt_net.core.types import (
@@ -102,7 +102,7 @@ class Plugin(BasePlugin):
                 )  # mode change
 
         elif name == Event.MODEL_BEFORE:
-            if "mode" in data and data["mode"] == MODE_VISION:
+            if data.get("mode") == MODE_CHAT:
                 key = self.get_option_value("model")
                 if self.window.core.models.has(key):
                     data['model'] = self.window.core.models.get(key)
@@ -119,7 +119,7 @@ class Plugin(BasePlugin):
                 data['value'] = self.on_system_prompt(data['value'])
 
         elif name == Event.UI_ATTACHMENTS:
-            mode = data["mode"]
+            mode = data.get("mode")
             if mode in [MODE_AGENT, MODE_AGENT_LLAMA, MODE_AGENT_OPENAI] and not self.window.core.config.get("cmd"):
                 pass
             else:
@@ -263,8 +263,7 @@ class Plugin(BasePlugin):
         # append vision prompt only if vision is provided or enabled
         if not self.is_vision_provided():
             return prompt
-        prompt = "Image attachment has been already sent.\n\n" + prompt
-        return prompt
+        return "Image attachment has been already sent.\n\n" + prompt
 
     def on_pre_prompt(self, prompt: str) -> str:
         """
@@ -294,12 +293,12 @@ class Plugin(BasePlugin):
         """
         mode = self.window.core.config.get('mode')
         attachments = self.window.core.attachments.get_all(mode)
-        self.window.core.gpt.vision.build_content(
+        self.window.core.api.openai.vision.build_content(
             str(self.prompt),
             attachments,
         )  # tmp build content, provide attachments from global mode
 
-        built_attachments = self.window.core.gpt.vision.attachments
+        built_attachments = self.window.core.api.openai.vision.attachments
         if len(built_attachments) > 0:
             return True
         return False
@@ -313,13 +312,13 @@ class Plugin(BasePlugin):
         result = False
         mode = self.window.core.config.get('mode')
         attachments = self.window.core.attachments.get_all(mode)  # from global mode
-        self.window.core.gpt.vision.build_content(
+        self.window.core.api.openai.vision.build_content(
             str(self.prompt),
             attachments,
         )  # tmp build content, provide attachments from global mode
 
-        built_attachments = self.window.core.gpt.vision.attachments
-        built_urls = self.window.core.gpt.vision.urls
+        built_attachments = self.window.core.api.openai.vision.attachments
+        built_urls = self.window.core.api.openai.vision.urls
 
         # check for images in URLs found in prompt
         img_urls = []
@@ -343,13 +342,13 @@ class Plugin(BasePlugin):
         :return: updated mode
         """
         # abort if already in vision mode or command enabled
-        if mode == MODE_VISION or mode in self.disabled_mode_switch:
+        if mode in self.disabled_mode_switch:
             return mode  # keep current mode
 
-        # if already used in this ctx then keep vision mode
+        # if already used in this ctx then keep vision (in CHAT) mode
         if self.is_vision_provided():
             ctx.is_vision = True
-            return MODE_VISION
+            return MODE_CHAT
 
         return mode  # keep current mode