fix: Pinned session TypeError, dead default models, get_history compat, i18n sync

- create_pinned_session built directly on aiohttp: the HA helper injects
  its own connector and raised TypeError on every setup and config flow
- session lifecycle: closed in APIClient.shutdown, on failed setup, and
  in config flow validators; allow_redirects=False everywhere so 3xx
  cannot route past the pinned resolver
- defaults: gemini-2.0-flash (shut down 2026-06-01) -> gemini-3.5-flash,
  deepseek-chat (retires 2026-07-24) -> deepseek-v4-flash
- DeepSeek V4 disable_thinking via thinking request parameter
- Gemini 3.x disable_thinking via thinking_level (MINIMAL, LOW for Pro)
- get_history limit: no forced default/max in schema, storage layer clamps
- translations: 6 locales caught up with strings.json, 2 orphan keys dropped
- README: manual install path matches zip layout, model sections updated

custom_components/ha_text_ai/const.py

@@ -58,8 +58,11 @@ MAX_HISTORY_FILE_SIZE = 1 * 1024 * 1024
 # Default values
 DEFAULT_MODEL: Final = "gpt-4o-mini"
 DEFAULT_ANTHROPIC_MODEL: Final = "claude-sonnet-4-6"
-DEFAULT_DEEPSEEK_MODEL: Final = "deepseek-chat"
-DEFAULT_GEMINI_MODEL: Final = "gemini-2.0-flash"
+# deepseek-chat/deepseek-reasoner are discontinued 2026-07-24; V4 models
+# select thinking mode via a request parameter instead of the model name.
+DEFAULT_DEEPSEEK_MODEL: Final = "deepseek-v4-flash"
+# gemini-2.0-flash was shut down 2026-06-01; 2.5-flash follows 2026-10-16.
+DEFAULT_GEMINI_MODEL: Final = "gemini-3.5-flash"
 DEFAULT_TEMPERATURE: Final = 0.1
 DEFAULT_MAX_TOKENS: Final = 1000
 DEFAULT_REQUEST_INTERVAL: Final = 1.0