feat(aws-bedrock/global.openai.gpt-5.6-sol): add new models [bot]

[models-bot]

Auto-generated by model-addition-agent for aws-bedrock/global.openai.gpt-5.6-sol.

---

Note

Low Risk
Declarative model catalog YAML only; no runtime, auth, or billing-code changes.

Overview
Adds a new AWS Bedrock model definition for global.openai.gpt-5.6-sol.

The catalog entry covers serverless chat/responses with thinking, function calling, prompt caching, and text+image input. Limits are a 1M context window and 128k max output. Per-region token costs include cache read/write and cumulative tiered pricing above 272k tokens, plus a reasoning_effort param (default medium).

^{Reviewed by Cursor Bugbot for commit 7c6ed90. Bugbot is set up for automated code reviews on this repo. Configure here.}

[github-actions]

/test-models

[github-actions]

/test-models

[cursor]

Cursor Bugbot has reviewed your changes using default effort and found 1 potential issue.

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

^{Reviewed by Cursor Bugbot for commit 7c6ed90. Configure here.}

[cursor]

Missing max_tokens param override

Medium Severity

limits now advertise max_tokens / max_output_tokens of 128000, but params only declares reasoning_effort. The model still inherits the provider default max_tokens maxValue of 2048 from default.yaml, so callers merging defaults with model params can reject valid requests up to the real 128k output limit. Comparable high-output Bedrock models and the native OpenAI gpt-5.6-sol entry override the token param maxValue to match their limits.

 

^{Reviewed by Cursor Bugbot for commit 7c6ed90. Configure here.}