feat(wafer/Kimi-K2.6): add new models [bot]

[models-bot]

Auto-generated by model-addition-agent for wafer/Kimi-K2.6.

---

Note

Low Risk
Additive provider metadata only; no runtime or auth logic changes.

Overview
Adds a new wafer provider entry for Kimi-K2.6 via providers/wafer/Kimi-K2.6.yaml.

The config defines token pricing (including cache-read input), a 262144 context window, function_calling, image input modality, and thinking: true. mode is set to unknown.

^{Reviewed by Cursor Bugbot for commit 988aafe. Bugbot is set up for automated code reviews on this repo. Configure here.}

[cursor]

Cursor Bugbot has reviewed your changes and found 2 potential issues.

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit 988aafe. Configure here.}

[cursor]

Chat model mode unknown

Medium Severity

The new wafer entry sets mode to unknown even though Kimi-K2.6 is a chat model with function_calling and thinking. Peer registry entries for the same model use mode: chat (and supportedModes: chat), so consumers that filter or route by mode may omit or mishandle this model.

 

^{Reviewed by Cursor Bugbot for commit 988aafe. Configure here.}

[cursor]

Cache cost without caching feature

Low Severity

costs includes cache_read_input_token_cost, but features does not list prompt_caching. Other Kimi-K2.6 configs and typical chat models in this repo pair cache-read pricing with prompt_caching, so capability flags and billing metadata can disagree for wafer.

 

^{Reviewed by Cursor Bugbot for commit 988aafe. Configure here.}