v4.5.0

In this update I tried to optimize and improve all the functionality of the bot, as well as unload heavy functions and divide all tasks into small functions.
I also began to study and design the Discord Embed functionality and that is why this version did not become 5.0.0 :)

Thanks for testing all code and functional: @Vasabiko

Added:

• Dynamic TTL caching
• Think and Reasoning tags now supported for send_split_message и send_split_code and use Discord Embed
• New AI models:
  • claude-3.7-sonnet
  • qwen-2.5-1m-demo
  • mixtral-small-24b

Improved/Fixed:

• pdfminer.six replaced by fitz (pyMuPDF). This lib more effective and stable
• Refactor message/command logic:
  • Removed process_message. Now we can use handle_response directly
  • handle_response is split into several functions
• Refactor Ban system:
  • New command /ban-info for check ban user status
  • Support Discord Embed
  • Now uses asyncio.task to automatically check for expired bans
• Refactor Remind system:
  • Improve time code logic
  • Now generate unique IDs
• Refactor Internet search:
  • Now function not create new session for each request. Use recall
  • Now support HTTP status errors
  • MAX_HTML_SIZE now Const
  • Add HTMLTooLarge checks
  • Pontential fix empty list result and unwanted element has early returned
• Refactor Libs checker:
  • Now uses importlib for effective checks and support more python versions

Remove:

• My RetryProvider class for evade duplicate logic. Now uses G4F RetryProvider

Full Changelog: v4.4.0...v4.5.0

v4.4.0

Sorry for long time update...

Added:

• Experimental internet-search configuration in internet_utils.py
• Checks for requierments.txt
• New models:
  o3-mini-low
  o3-mini
  Gemini 2.0 Flash
  Gemini 2.0 Thinking

Improved:

• New internet-search algorithm
• Response messages are more clean

Updated:

• G4F to 0.4.6.2
• Cryptography to 44.0.1
• duckduckgo_search tp 7.4.2
• beautifulsoup4 to 4.13.3

Removed:

• /draw-prodia service now not avalable with api v1

Full Changelog: v4.3.0...v4.4.0

v4.3.0

v4.3.0

Contributors:

Thanks for testing: @Vasabiko
Thanks for improve g4f for ideas and corrections that were implemented: @kqlio67

Added:

• Encryption user_data support
• Individual instructions for each users
  • Use commands:
    /instuction-set
/instuction-reset
• Multi image generation (1-4 images)
• EXPERIMENTAL! Add api_keys support for auth providers (WIP)

  If you are an experienced developer and can help improve this implementation,
  I would be extremely grateful!

• New model:
  Deepseek-R1

Improved:

• More providers for some models

Fixed:

• Some missing imports

Updated:

• G4F to 0.4.2.6

Removed:

• Some old AI models cause they now not popular

Full Changelog: v.4.2.0...v4.3.0

v4.2.0-hotfix

v4.2.0-hotfix

Hotfix:

• Added missing locale folder for fix startup bot

Full Changelog: v.4.2.0...v4.2.0-hotfix

v.4.2.0

v4.2.0

Added:

• .har/cookies G4F support
• Mention support
• New AI models (Auth providers):
  1. Command R+ и R7B+
  2. Nvidia Nemotron 70B
  3. Qwen 2/2.5/Coder/Vision 72B/32B
  4. Cably 80B
  5. Phi 3.5 Mini
  6. GLM-4 230B

Improved:

• Semi-Refactoring code for optimize and more compatibility
• Prepare to locale support
• Web Search instruction logic for each param
• Banned user check logic
• Image drawing logic
• Read/Write files logic
• Code marking after split message chunks

Fixed:

• Ban/Unban functional

Updated:

• G4F lib for 0.4.1.1

Full Changelog: v4.1.0...v.4.2.0

v4.1.0

v4.1.0

Added Admin functional with commands:

Request set ADMIN_ID in .env file

• /ban
• /unban
• /banned-list

Improved:

• Bot startup checks by @kqlio67
• .env params checks by @kqlio67
• Validate channels by @kqlio67
• Enhance error handlings by @kqlio67 and Me
• Web search functional by @kqlio67 and Me
• Optimized .gitignore and requirements.txt by @kqlio67

Fixed

• Web search after duckduckgo_search update

Pull requests

• Update providers, enhance error handling and optimize dependencies by @kqlio67 in #4

New Contributors

• @kqlio67 made their first contribution in #4

Full Changelog: v4.0.0...v4.1.0

v4.0.0

v4.0.0

NEW MAIN:

• Now have Remind/Notification functional.

Added:

• /draw-prodia command for image generation with Prodia provider
• /remind-add - add new remind for user
• /remind-list - show your remind list
• /remind-delete - delete your remind (index)

Fixed:

• Image response for discord message

Updated:

• tests.py have more friendly config params
• g4f.stubs updated after G4F update. Thanks for note @kqlio67

v3.5.0 HotFix

Fix syntax error for enqueue_message (Sorry)

Full Changelog: v3.5.0...v3.5.0-hotfix

v3.5.0

• Updated G4F to 0.3.8.0
• Add new command: /askpdf
  • Now bot support PDF reading (Text only. Images does not see)
• Improve Web search functionality (Use gather)
• Improve Stability bot:
  • Use mimetypes for /asklong and /askpdf commands for correctly check .ext files
  • Add more checks for missing .env params and set some params to default
  • Add G4F_DEBUG .env param for use G4F debug functions and logging.
  • Set gpt-4o to default
  • Providers and models list updated to latest
• Remove /chat-model-uncensored
  • Templorary removing cause G4F Airforce provider problem
• Update tests.py:
  • Now should normaly work TimeOut checks
  • Now support param: parallel_execution (true/false)

v3.4.0

• Added /chat-model-uncensored command
• .env now have ADMIN_NAME and BOT_NAME
• .env now use gpt-4o model to default
• Refactor default provider code for model. Now take automatic.
• Fixed error loop then all providers for model not working
• Updated requirements and add check for 3.13+ python
• Updated for G4F 0.3.3.4 (Not i didnt want use 0.3.4.0 for some problems. This version more stable).
• Removed /model-info command