Confirm qBraid Azure quota tier for gpt-5.6 + file prod quota-increase request

[Kenny-Heitritter]

Question

Confirm the qbraid-prod Azure subscription's assigned quota tier for gpt-5.6-sol/terra/luna (GlobalStandard, eastus2), and file a quota-increase request if it is below Tier 5–6. This fixes the achievable "much higher" prod rate limit that #34 deploys against.

Live-only (requires the qBraid Azure portal / control-plane API), spun out of #32 per that ticket's instruction. Everything not live is already doc-confirmed in #32's resolution:

• eastus2 × GlobalStandard is available for all three models (v2026-07-09).
• Per-tier GlobalStandard ceilings (TPM = sku_capacity × 1000): Tier 1 = 1M … Tier 5 = 10M … Tier 6 = 15M. Tier 5/6 have gpt-5.6 quota by default.

Deliverable: the subscription's current quota tier + the sku_capacity (TPM) prod should target (recommend 10,000–15,000 = 10–15M TPM), and — if the subscription is below that — a filed quota-increase request (https://aka.ms/oai/stuquotarequest).

Check tier via the control-plane API:
GET https://management.azure.com/subscriptions/<sub>/providers/Microsoft.CognitiveServices/quotaTiers?api-version=2025-10-01-preview

Likely needs the human's Azure access (HITL) unless az is authenticated for the prod subscription.

Part of #31.

[Kenny-Heitritter]

We can get up to 10M tok/s on our current tier.