feat(azure-open-ai/gpt-5.6-luna-2026-07-09): add new models [bot]

[models-bot]

Auto-generated by model-addition-agent for azure-open-ai/gpt-5.6-luna-2026-07-09.

---

Note

Low Risk
Metadata-only addition with no runtime or auth changes; main gap is missing cost data until Azure publishes pricing.

Overview
Registers gpt-5.6-luna-2026-07-09 in the Azure Open AI provider via a new catalog YAML (providers/azure-open-ai/gpt-5.6-luna-2026-07-09.yaml), auto-generated by the model-addition agent.

The entry marks the model active with serverless provisioning, chat mode, thinking enabled, text/image input, and standard tool/structured-output features. Token limits are set to a ~1.05M context window (922k input / 128k output), with configurable reasoning_effort (default medium, including none through xhigh). Retirement is set to 2027-07-09.

Pricing/costs are intentionally omitted with a note that Azure authoritative pricing is not published yet for this model.

^{Reviewed by Cursor Bugbot for commit c15baea. Bugbot is set up for automated code reviews on this repo. Configure here.}

[github-actions]

/test-models

[harshiv-26]

Gateway test results

• Total: 0
• Passed: 0
• Failed: 0
• Validation failed: 0
• Errored: 0
• Skipped: 0
• Success rate: 0.0%

Provider	Model	Scenarios

[cursor]

Cursor Bugbot has reviewed your changes using default effort 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 c15baea. Configure here.}

[cursor]

Missing responses supported mode

Medium Severity

supportedModes lists only chat, while the OpenAI gpt-5.6-luna entry and nearby Azure GPT‑5.x configs (e.g. gpt-5.5-2026-04-24) also include responses. Consumers that gate the Responses API on supportedModes may treat this model as chat-only incorrectly.

 

^{Reviewed by Cursor Bugbot for commit c15baea. Configure here.}

[cursor]

max_tokens cap contradicts limits

Medium Severity

limits sets max_tokens and max_output_tokens to 128000, but params does not override max_tokens. Azure default.yaml still caps max_tokens at 4096, so merged config can advertise a 128k output window while validating requests against 4096.

 

^{Reviewed by Cursor Bugbot for commit c15baea. Configure here.}