[2026.2.13.2] - 2026-02-13

📝 Changes since v2026.2.13.1

📦 Changes

• 🔖 Bump version to 2026.2.13.2 (ce34056)
• Merge branch 'main' of https://github.com/OEvortex/Webscout (bc449d9)
• Enhance TypliAI and Toolbaz providers with CSRF token handling, support for reasoning models, and improved streaming response management. Update changelog with fixes and new features. (f6f6a57)

📦 Installation

pip install --upgrade webscout==2026.2.13.2

[2026.2.13.1] - 2026-02-13

📝 Changes since v2026.2.13

📦 Changes

• 🔖 Bump version to 2026.2.13.1 (2f14266)
• Refactor code structure for improved readability and maintainability (5e9e511)

📦 Installation

pip install --upgrade webscout==2026.2.13.1

[2026.2.13] - 2026-02-13

📝 Changes since v2026.2.11

📦 Changes

• 🔖 Bump version to 2026.2.13 (c6460bc)
• Merge branch 'main' of https://github.com/OEvortex/Webscout (4da8e62)
• Refactor providers to use curl_cffi for HTTP requests and add Falcon provider (a9d8088)

📦 Installation

pip install --upgrade webscout==2026.2.13

[2026.2.11] - 2026-02-11

📝 Changes since v2026.1.31

📦 Changes

• 🔖 Bump version to 2026.2.11 (e3accb8)
• feat: Consolidate LitAgent imports and remove redundant lines in deepinfra.py and elmo.py (67172e8)
• Refactor Provider imports, remove X0GPT, and add K2Think integration (dcd2386)

📦 Installation

pip install --upgrade webscout==2026.2.11

[2026.1.31] - 2026-01-31

📝 Changes since v2026.1.22

📦 Changes

• 🔖 Bump version to 2026.1.31 (da5af4b)
• Merge pull request #114 from OEvortex/new-providers (cb36b3c)
• feat: Improve type safety and error handling across provider implementations (eddefc9)
• feat: Enhance ProxyManager with improved error handling and response validation (b0871e4)
• feat: Add ProxyManager utility for managing and rotating proxies with auto-fetching support refactor: Update model fetcher cache to use system temp directory and add debug logging chore: Remove unused lol.py file docs: Update changelog with new ProxyManager details (e690925)
• Refactor model fetching across providers to use non-blocking background fetch (53cc9d5)
• feat: Add Upstage provider with OpenAI-compatible interface and implement streaming support fix: Update provider documentation and changelog for new Upstage integration test: Add unit tests for Upstage provider functionality and response handling refactor: Rename OPENAI directory to Openai_comp for consistency (d26d8c2)
• Implement non-blocking background model fetching with caching support (6249adb)
• feat: Update DeepInfra provider by removing deprecated model list and enhancing task orchestration instructions (6fdaaaa)
• Add Writecream, X0GPT, and Zenmux providers with chat completion functionality (e237371)
• feat: Enhance Kyutai TTS provider with new models and comprehensive tests (8519523)
• feat: Add Pocket TTS provider implementation and include it in the TTS module (0230cb6)
• feat: Add new grammar checker and AI humanizer provider implementations (fd19823)
• feat: Move BraveAI provider to UNFINISHED folder; add AyeSoul provider implementation with streaming support and image upload functionality (7502445)

📦 Installation

pip install --upgrade webscout==2026.1.31

[2026.1.22] - 2026-01-22

📝 Changes since v2026.1.19

📦 Changes

• 🔖 Bump version to 2026.1.22 (5a78352)
• Merge branch 'main' of https://github.com/pyscout/Webscout (cbeda58)
• feat: Remove HadadXYZ provider implementations and enhance DeepInfra model filtering; add tests for OpenAI DeepInfra provider (dac1494)
• chore: remove HadadXYZ provider implementations (97b6846)

📦 Installation

pip install --upgrade webscout==2026.1.22

[2026.1.19] - 2026-01-19

📝 Changes since v2026.1.18

📦 Changes

• 🔖 Bump version to 2026.1.19 (acfc3ce)
• feat: Update Ayle provider models and remove MiroMind integration; add tests for Ayle models (055b53f)

📦 Installation

pip install --upgrade webscout==2026.1.19

[2026.1.18] - 2026-01-18

📝 Changes since v2026.1.17

📦 Changes

• 🔖 Bump version to 2026.1.18 (90a0280)
• Delete uv.lock (478ded0)

📦 Installation

pip install --upgrade webscout==2026.1.18

[2026.1.17] - 2026-01-17

📝 Changes since v2026.1.16

[2026.01.17] - 2026-01-17

🔧 Improved

• refactor: webscout/Provider/AISEARCH/iask_search.py - Streamlined IAsk AI search provider:
  • Removed Phoenix WebSocket complexity and unused cache_find() helper function
  • Simplified API flow to single GET request to /q endpoint with mode and detail_level parameters
  • Added _extract_answer() method to parse HTML response and extract answer from div#id="text"
  • Added _iter_chunks() method for improved chunked streaming (800-char chunks with word boundaries)
  • Removed unnecessary aiohttp WebSocket dependency, simplified async implementation
  • Updated endpoints: api_endpoint to https://iask.ai, new query_endpoint to https://iask.ai/q
  • Integrated LitAgent for realistic User-Agent header generation (replaced hardcoded static user agent)
  • Both streaming and non-streaming modes now use consistent fetch_answer() async flow
  • Performance: Reduced request overhead and latency by eliminating WebSocket handshake

🐛 Fixed

� Fixed

• fix: webscout/Provider/OPENAI/sonus.py - Fixed critical non-stream request error "stream mode is not enabled":
  • Root cause: SonusAI API returns a streaming response even for non-stream requests
  • Solution: Added stream=True parameter to the session.post() call in _create_non_stream() method
  • Impact: Non-streaming chat completions now work correctly with the OpenAI-compatible SonusAI provider

📝 Documentation

• docs: webscout/client.py - Generated comprehensive PEP 257-compliant docstrings for all 32+ classes and methods:
  • Added detailed module-level docstrings for load_openai_providers(), load_tti_providers(), and _get_models_safely()
  • Documented ClientCompletions class with intelligent provider/model resolution algorithms
  • Documented ClientChat adapter interface for chat completions
  • Documented ClientImages class with automatic failover for image generation
  • Documented main Client class with comprehensive usage examples
  • All docstrings include parameter descriptions, return types, exception documentation, usage examples, and important behavioral notes
  • Improved IDE autocomplete and automated documentation generation support

🔧 Improved

• refactor: webscout/Provider/akashgpt.py - Replaced cloudscraper with curl_cffi for better performance and TLS/SSL handling. Changed from cloudscraper.create_scraper() to curl_cffi.requests.Session() for HTTP requests.
• refactor: webscout/Provider/OPENAI/akashgpt.py - Replaced cloudscraper with curl_cffi for improved HTTP request handling. Updated session initialization to use curl_cffi.requests.Session() with proper proxy support in the __init__ method.

✅ Quality

• All ruff linting checks passed successfully on all modified files
• All type checking with ty passed successfully

📦 Installation

pip install --upgrade webscout==2026.1.17

Brave Search

📝 Changes since v2026.1.6

📦 Changes

• 🔖 Bump version to 2026.1.16 (6793ed0)
• Merge pull request #113 from OEvortex/feature/search (c6bde57)
• feat: Remove deprecated testing script for Brave search engines (8896493)
• feat: Enhance Brave search functionality with new engines and improved parsing (56775b4)
• feat: Implement BraveImages search engine and associated tests; refactor Brave engine; update documentation and usage instructions (251938f)
• feat: Add agent orchestration and subagent definitions for enhanced AI capabilities; remove unused ChatZAI provider and line counter script; introduce docstring and reverse engineering agents (655c285)
• Jan 15, 2026, 3:09 PM (0dbcb78)
• Jan 15, 2026, 3:08 PM (05dd54a)
• Jan 15, 2026, 2:59 PM (3e394b2)
• Jan 15, 2026, 2:57 PM (27f7e58)
• feat: Update BraveAI provider to support additional parameters for language and localization (c8599d1)
• feat: Enhance BraveAI provider with additional parameters for research and localization (1d41add)
• feat: Add BraveAI provider for AI-generated responses using Brave Search (17b8f04)

📦 Installation

pip install --upgrade webscout==2026.1.16

What's Changed

• [2026.01.16] - 2026-01-16 by @OEvortex in #113

Full Changelog: v2026.1.6...v2026.1.16