feat(aihubmix): add Qwen3.7 Max, Gemini 3.5 Flash, Grok 4.3 models and improve Anthropic support

[orangeclk]

Summary

New Models

• Qwen3.7 Max — Supports deep thinking (enable_thinking), tool call / stream-tool-call / multi-tool-call, agent-thought, ~1M context window, with pricing configured.
• Gemini 3.5 Flash
• Grok 4.3

Anthropic Models Enhancement

• Added default max_tokens parameter for Anthropic models to prevent request failures caused by missing value.

Change Type

• Documentation / non-plugin change
• Non-LLM plugin (tools, extensions, datasource, etc.)
• LLM plugin

Screenshots / Videos

Before	After

LLM Plugin Checklist

Areas affected by this change (check all that apply)

• Message flow (system messages, user ↔ assistant turn-taking)
• Tool interaction flow (multi-round usage, Agent App and Agent Node)
• Multimodal input (images, PDFs, audio, video, etc.)
• Multimodal output (images, audio, video, etc.)
• Structured output (JSON, XML, etc.)
• Token consumption metrics
• Other LLM functionality (reasoning, grounding, prompt caching, etc.)
• New models / model parameter fixes

Version

• Bumped top-level version in manifest.yaml (not the one under meta)
• dify_plugin>=0.3.0,<0.6.0 is declared in pyproject.toml and locked in uv.lock (or kept in requirements.txt for legacy plugins without uv.lock) — SDK docs

Testing

• Local deployment — Dify version: 1.14.2
• SaaS (cloud.dify.ai)

[gemini-code-assist]

Code Review

This pull request adds support for several new models, including Qwen 3.7 Max, Gemini 3.5 Flash, and Grok 4.3, and updates the Anthropic model implementation. Feedback identifies a potential bug in the Anthropic integration where a high default max_tokens value could cause API errors. For the Gemini 3.5 Flash configuration, the reviewer noted a missing response_format parameter needed for structured outputs and a discrepancy between the defined default value and the help text for the thinking_level parameter.