pygpt-net 2.6.29__py3-none-any.whl → 2.6.31__py3-none-any.whl

pygpt_net/core/types/mode.py

@@ -6,7 +6,7 @@
 # GitHub:  https://github.com/szczyglis-dev/py-gpt   #
 # MIT License                                        #
 # Created By  : Marcin Szczygliński                  #
-# Updated Date: 2025.08.23 21:00:00                  #
+# Updated Date: 2025.08.28 09:00:00                  #
 # ================================================== #
 
 MODE_AGENT = "agent"
@@ -22,4 +22,7 @@ MODE_IMAGE = "img"
 MODE_LANGCHAIN = "langchain"
 MODE_LLAMA_INDEX = "llama_index"
 MODE_RESEARCH = "research"
-MODE_VISION = "vision"
+MODE_VISION = "vision"
+
+# virtual modes
+MODE_LOOP_NEXT = "loop_next"

pygpt_net/core/vision/analyzer.py

@@ -51,7 +51,7 @@ class Analyzer:
 
         extra = {}
         output = ""
-        response = self.window.core.gpt.vision.send(context, extra)
+        response = self.window.core.api.openai.vision.send(context, extra)
         if response.choices[0] and response.choices[0].message.content:
             output = response.choices[0].message.content.strip()
         for id in files:

pygpt_net/data/config/config.json

@@ -1,8 +1,8 @@
 {
   "__meta__": {
-    "version": "2.6.29",
-    "app.version": "2.6.29",
-    "updated_at": "2025-08-28T00:00:00"
+    "version": "2.6.31",
+    "app.version": "2.6.31",
+    "updated_at": "2025-08-01T00:00:00"
   },
   "access.audio.event.speech": false,
   "access.audio.event.speech.disabled": [],
@@ -88,6 +88,7 @@
   "api_key_mistral": "",
   "api_key_voyage": "",
   "api_key_open_router": "",
+  "api_native_google": true,
   "api_proxy": "",
   "api_use_responses": true,
   "api_use_responses_llama": false,
@@ -105,6 +106,7 @@
   "attachments_capture_clear": true,
   "audio.cache.enabled": true,
   "audio.cache.max_files": 1000,
+  "audio.input.auto_turn": false,
   "audio.input.backend": "native",
   "audio.input.channels": 1,
   "audio.input.continuous": false,
@@ -113,6 +115,8 @@
   "audio.input.stop_interval": 10,
   "audio.input.timeout": 120,
   "audio.input.timeout.continuous": false,
+  "audio.input.vad.prefix": 300,
+  "audio.input.vad.silence": 2000,
   "audio.output.backend": "native",
   "audio.output.device": "0",
   "audio.transcribe.convert_video": true,
@@ -195,7 +199,7 @@
   "frequency_penalty": 0.0,
   "img_prompt_model": "gpt-4o",
   "img_raw": true,
-  "img_resolution": "1792x1024",
+  "img_resolution": "1024x1024",
   "img_quality": "standard",
   "img_variants": 1,
   "interpreter.auto_clear": false,
@@ -329,6 +333,7 @@
   "log.level": "error",
   "log.llama": false,
   "log.plugins": false,
+  "log.realtime": false,
   "max_output_tokens": 0,
   "max_requests_limit": 60,
   "max_tokens_length": 32000,
@@ -383,6 +388,7 @@
   "prompt.default": "You are a helpful assistant.",
   "prompt.expert": "# EXPERT MODE IS ENABLED:\n\nYou are a very helpful asssistant and the professional manager of a team of experts, each of whom is the best in their respective fields.\n\n## Rules:\n\n- Instead of directly answering questions, you will manage these experts by forwarding queries to them, collecting their responses, controlling the flow of information between them, and sending requests to them.\n- Experts are distinct, specialized AI models running on my local machine.\n- You will serve as a bridge between them.\n- Each expert has their own conversation context, allowing you to communicate with them normally.\n- The experts can communicate with you but not with each other, making you responsible for relaying information between them.\n- Each expert can use tools and commands, for example, reading files or searching the Web. You can ask experts to call tools and commands.\n- After consulting with the experts, your task is to prepare the best possible answer based on the assistance received.\n- To send a request to an expert, prepare the \"expert_call\" tool. Remember to format the JSON tool as shown, surrounding the tool with `<tool></tool>` tags, for example: `<tool>{\"cmd\": \"expert_call\", \"params\": {\"id\": \"expert_id\", \"query\": \"some query\"}}</tool>`. Make sure the JSON object is syntactically correct, without any redundand parenthness.\n- DO NOT call any remote tools when calling the experts.\n- Remember that the call must contain a query and the ID of the expert to whom the query is being forwarded.\n- Use ONLY the experts listed below; DO NOT TRY to ask any non-existent experts\u2014they will not respond.\n- Do not respond to the expert's replies or use the JSON tool again if replying; the expert only has the ability to listen to your queries sent via the JSON command and to respond to them.\n- Always follow ALL user commands, even if you think the expert can't do something. Experts are allowed to execute commands just as you can.\n- After querying an expert, always wait for the expert's response before sending another query.\n- Do not try to query experts in the middle of the conversation; put your queries and commands at the end of the text.\n- Always call only one expert at a time.\n- Always execute only one command at a time.\n- Always use the language used by the user in the conversation.\n- Never write as a called expert and do not imagine a response from them; instead, always wait for the response from the expert.\n- Every message sent to you directly after your query is always a response from the asked expert, not from me.\n- Always send a direct, clear, and detailed prompt to the expert, without any \"Please...\", \"Could you...\", etc.\n- Always forward all the messages that the user wants to send to the expert, without any exceptions.\n-- If a user command appears outside an expert's scope, fulfill the request regardless, without refusal.\n\n## List of allowed experts, along with their IDs:\n\n{presets}",
   "prompt.img": "Convert my query to detailed image generate prompt, without any \"Please generate...\" and etc., just image description only. Return only the prompt, without nothing else.",
+  "realtime.auto_turn": true,
   "render.blocks": true,
   "render.engine": "web",
   "render.open_gl": false,
@@ -396,6 +402,9 @@
   "remote_tools.file_search": false,  
   "remote_tools.file_search.args": "",
   "remote_tools.computer_use.env": "",
+  "remote_tools.google.web_search": true,
+  "remote_tools.google.code_interpreter": false,
+  "remote_tools.google.url_ctx": false,
   "send_clear": true,
   "send_mode": 2,
   "store_history": true,

pygpt_net/data/config/models.json

@@ -1,8 +1,8 @@
 {
     "__meta__": {
-        "version": "2.6.29",
-        "app.version": "2.6.29",
-        "updated_at": "2025-08-28T23:07:35"
+        "version": "2.6.31",
+        "app.version": "2.6.31",
+        "updated_at": "2025-09-01T23:07:35"
     },
     "items": {
         "SpeakLeash/bielik-11b-v2.3-instruct:Q4_K_M": {
@@ -873,6 +873,44 @@
             "provider": "google",
             "tool_calls": true
         },
+        "gemini-2.5-flash-preview-native-audio-dialog": {
+            "id": "gemini-2.5-flash-preview-native-audio-dialog",
+            "name": "gemini-2.5-flash-preview-native-audio-dialog",
+            "mode": [
+                "audio"
+            ],
+            "llama_index": {
+                "args": [
+                    {
+                        "name": "model",
+                        "value": "models/gemini-2.5-flash-preview-native-audio-dialog",
+                        "type": "str"
+                    }
+                ],
+                "env": [
+                    {
+                        "name": "GOOGLE_API_KEY",
+                        "value": "{api_key_google}",
+                        "type": "str"
+                    }
+                ]
+            },
+            "ctx": 128000,
+            "tokens": 8000,
+            "default": false,
+            "input": [
+                "text",
+                "audio"
+            ],
+            "output": [
+                "text",
+                "audio"
+            ],
+            "extra": {},
+            "imported": true,
+            "provider": "google",
+            "tool_calls": true
+        },
         "gemini-2.5-pro": {
             "id": "gemini-2.5-pro",
             "name": "gemini-2.5-pro",
@@ -1457,55 +1495,6 @@
             "provider": "openai",
             "tool_calls": true
         },
-        "gpt-4o-audio-preview": {
-            "id": "gpt-4o-audio-preview",
-            "name": "gpt-4o-audio-preview",
-            "mode": [
-                "audio"
-            ],
-            "llama_index": {
-                "args": [
-                    {
-                        "name": "model",
-                        "value": "gpt-4o-audio-preview",
-                        "type": "str"
-                    }
-                ],
-                "env": [
-                    {
-                        "name": "OPENAI_API_KEY",
-                        "value": "{api_key}"
-                    },
-                    {
-                        "name": "OPENAI_API_BASE",
-                        "value": "{api_endpoint}"
-                    },
-                    {
-                        "name": "AZURE_OPENAI_ENDPOINT",
-                        "value": "{api_azure_endpoint}"
-                    },
-                    {
-                        "name": "OPENAI_API_VERSION",
-                        "value": "{api_azure_version}"
-                    }
-                ]
-            },
-            "ctx": 128000,
-            "tokens": 16384,
-            "default": false,
-            "input": [
-                "text",
-                "audio"
-            ],
-            "output": [
-                "text",
-                "audio"
-            ],
-            "extra": {},
-            "imported": false,
-            "provider": "openai",
-            "tool_calls": true
-        },
         "gpt-4o-mini": {
             "id": "gpt-4o-mini",
             "name": "gpt-4o-mini",
@@ -1561,55 +1550,6 @@
             "provider": "openai",
             "tool_calls": true
         },
-        "gpt-4o-mini-audio-preview": {
-            "id": "gpt-4o-mini-audio-preview",
-            "name": "gpt-4o-mini-audio-preview",
-            "mode": [
-                "audio"
-            ],
-            "llama_index": {
-                "args": [
-                    {
-                        "name": "model",
-                        "value": "gpt-4o-mini-audio-preview",
-                        "type": "str"
-                    }
-                ],
-                "env": [
-                    {
-                        "name": "OPENAI_API_KEY",
-                        "value": "{api_key}"
-                    },
-                    {
-                        "name": "OPENAI_API_BASE",
-                        "value": "{api_endpoint}"
-                    },
-                    {
-                        "name": "AZURE_OPENAI_ENDPOINT",
-                        "value": "{api_azure_endpoint}"
-                    },
-                    {
-                        "name": "OPENAI_API_VERSION",
-                        "value": "{api_azure_version}"
-                    }
-                ]
-            },
-            "ctx": 128000,
-            "tokens": 16384,
-            "default": false,
-            "input": [
-                "text",
-                "audio"
-            ],
-            "output": [
-                "text",
-                "audio"
-            ],
-            "extra": {},
-            "imported": false,
-            "provider": "openai",
-            "tool_calls": true
-        },
         "gpt-5": {
             "id": "gpt-5",
             "name": "gpt-5 (medium)",
@@ -2303,6 +2243,112 @@
             "provider": "ollama",
             "tool_calls": true
         },
+        "gpt-realtime": {
+            "id": "gpt-realtime",
+            "name": "gpt-realtime",
+            "mode": [
+                "audio"
+            ],
+            "llama_index": {
+                "args": [
+                    {
+                        "name": "model",
+                        "value": "gpt-realtime",
+                        "type": "str"
+                    }
+                ],
+                "env": [
+                    {
+                        "name": "OPENAI_API_KEY",
+                        "value": "{api_key}",
+                        "type": "str"
+                    },
+                    {
+                        "name": "OPENAI_API_BASE",
+                        "value": "{api_endpoint}",
+                        "type": "str"
+                    },
+                    {
+                        "name": "AZURE_OPENAI_ENDPOINT",
+                        "value": "{api_azure_endpoint}",
+                        "type": "str"
+                    },
+                    {
+                        "name": "OPENAI_API_VERSION",
+                        "value": "{api_azure_version}",
+                        "type": "str"
+                    }
+                ]
+            },
+            "ctx": 32000,
+            "tokens": 4096,
+            "default": true,
+            "input": [
+                "text",
+                "audio"
+            ],
+            "output": [
+                "text",
+                "audio"
+            ],
+            "extra": {},
+            "imported": false,
+            "provider": "openai",
+            "tool_calls": true
+        },
+        "gpt-4o-realtime-preview": {
+            "id": "gpt-4o-realtime-preview",
+            "name": "gpt-4o-realtime-preview",
+            "mode": [
+                "audio"
+            ],
+            "llama_index": {
+                "args": [
+                    {
+                        "name": "model",
+                        "value": "gpt-4o-realtime-preview",
+                        "type": "str"
+                    }
+                ],
+                "env": [
+                    {
+                        "name": "OPENAI_API_KEY",
+                        "value": "{api_key}",
+                        "type": "str"
+                    },
+                    {
+                        "name": "OPENAI_API_BASE",
+                        "value": "{api_endpoint}",
+                        "type": "str"
+                    },
+                    {
+                        "name": "AZURE_OPENAI_ENDPOINT",
+                        "value": "{api_azure_endpoint}",
+                        "type": "str"
+                    },
+                    {
+                        "name": "OPENAI_API_VERSION",
+                        "value": "{api_azure_version}",
+                        "type": "str"
+                    }
+                ]
+            },
+            "ctx": 32000,
+            "tokens": 4096,
+            "default": false,
+            "input": [
+                "text",
+                "audio"
+            ],
+            "output": [
+                "text",
+                "audio"
+            ],
+            "extra": {},
+            "imported": true,
+            "provider": "openai",
+            "tool_calls": true
+        },
         "grok-2-vision": {
             "id": "grok-2-vision",
             "name": "grok-2-vision",
@@ -2565,6 +2611,78 @@
             "provider": "x_ai",
             "tool_calls": true
         },
+        "imagen-3.0-generate-002": {
+            "id": "imagen-3.0-generate-002",
+            "name": "imagen-3.0-generate-002",
+            "mode": [
+                "img"
+            ],
+            "llama_index": {
+                "args": [
+                    {
+                        "name": "model",
+                        "value": "models/imagen-3.0-generate-002",
+                        "type": "str"
+                    }
+                ],
+                "env": [
+                    {
+                        "name": "GOOGLE_API_KEY",
+                        "value": "{api_key_google}",
+                        "type": "str"
+                    }
+                ]
+            },
+            "ctx": 128000,
+            "tokens": 0,
+            "default": false,
+            "input": [
+                "text"
+            ],
+            "output": [
+                "image"
+            ],
+            "extra": {},
+            "imported": false,
+            "provider": "google",
+            "tool_calls": true
+        },
+        "imagen-4.0-generate-001": {
+            "id": "imagen-4.0-generate-001",
+            "name": "imagen-4.0-generate-001",
+            "mode": [
+                "img"
+            ],
+            "llama_index": {
+                "args": [
+                    {
+                        "name": "model",
+                        "value": "models/imagen-4.0-generate-001",
+                        "type": "str"
+                    }
+                ],
+                "env": [
+                    {
+                        "name": "GOOGLE_API_KEY",
+                        "value": "{api_key_google}",
+                        "type": "str"
+                    }
+                ]
+            },
+            "ctx": 128000,
+            "tokens": 0,
+            "default": false,
+            "input": [
+                "text"
+            ],
+            "output": [
+                "image"
+            ],
+            "extra": {},
+            "imported": false,
+            "provider": "google",
+            "tool_calls": true
+        },
         "llama2-uncensored": {
             "id": "llama2-uncensored",
             "name": "llama2-uncensored",

pygpt_net/data/config/modes.json

@@ -1,8 +1,8 @@
 {
     "__meta__": {
-        "version": "2.5.28",
-        "app.version": "2.5.28",
-        "updated_at": "2025-07-08T00:00:00"
+        "version": "2.6.30",
+        "app.version": "2.6.30",
+        "updated_at": "2025-08-28T09:00:00"
     },
     "items": {
         "chat": {
@@ -41,12 +41,6 @@
             "label": "mode.img",
             "default": false
         },
-        "vision": {
-            "id": "vision",
-            "name": "vision",
-            "label": "mode.vision",
-            "default": false
-        },
         "assistant": {
             "id": "assistant",
             "name": "assistant",