feat(nano-gpt): add Kimi K2.6 and Qwen 3.6 models

[sk0x0y]

Summary

• Add Kimi K2.6 (moonshotai/kimi-k2.6) and Kimi K2.6 Thinking (moonshotai/kimi-k2.6:thinking) to the NanoGPT provider
• Add Qwen 3.6 model family to the NanoGPT provider:
  • qwen-3.6-plus — commercial multimodal agent model, ~1M context
  • qwen3.6-max-preview — flagship Qwen 3.6 model
  • qwen/Qwen3.6-35B-A3B — open-weights MoE vision-language model
  • qwen/Qwen3.6-35B-A3B:thinking — reasoning variant of the 35B A3B
  • alibaba/qwen3.6-flash — fast vision-language model, ~1M context

All model metadata (pricing, context limits, modalities, capabilities) sourced from nano-gpt.com model pages.

Validation

• bun validate passes with exit code 0