This release expands VTX Macro from the previous 7-provider cloud setup to a unified 16-provider BYOK experience.

The newly added first-class BYOK providers are:

• Alibaba Cloud Model Studio
• Amazon Bedrock
• Cerebras
• DeepSeek
• Fireworks AI
• Lightning AI
• Mistral
• Moonshot/Kimi
• Together AI

They join the existing BYOK providers: Anthropic, Google Gemini, Groq, OpenAI, OpenRouter, Venice AI, and xAI.

The model catalog now reports 859 available AI models, giving users a much wider range of choices for trading, reviews, and insights. A new Models page makes that catalog easier to explore with search, provider filtering, available-key filtering, context-window sorting, price sorting, and model/provider comparisons.

Free-model discovery is also clearer. VTX Macro now identifies 33 confirmed free models in the current catalog, and the model selectors include a Free filter so users can quickly find models with confirmed $0.00 input and output token pricing. Unknown prices are no longer treated as free: if VTX Macro cannot safely confirm a token price, the UI shows it as unavailable instead of forcing it to $0.00.

AI settings also gain more transparent reasoning and thinking controls. Supported models expose provider-native options such as effort levels, thinking budgets, or thinking toggles, while leaving a control unselected sends no override and lets the provider default apply.

Together, this release makes BYOK model choice broader, pricing easier to trust, and provider behavior more visible before users put a model into live trading or analysis.