Azure regional changes — 2026-08-12

· Auto summary
383 new196 regressions13 parked
Daily executive summary

Compared with the 2026-08-11 snapshot, the 2026-08-12 scan recorded 383 new listings, 196 regressions, and 13 parked unknown transitions.

Where it changed: VM SKUs: 296 new listings, 105 regressions; Azure AI models: 28 new listings, 91 regressions; AKS extensions: 59 new listings, 0 regressions.

Representative changes: cohere command r 08 2024 model from cohere (version 1) was no longer listed in 31 regions; cohere command r plus 08 2024 model from cohere (version 1) was no longer listed in 31 regions; mai thinking 1 model from Microsoft (version 2026-06-01) was no longer listed in 29 regions.

What changed: 92 first recorded disappearances, 104 recurring disappearances, 383 net-new regional listings.

What this means: Review the 196 regressions against existing regional targets and fallbacks first; treat the 383 new listings as options to validate. These are read-only catalog/list signals, not quota, capacity, or deployment results. Zero new delistings does not establish recovery of earlier missing listings.

7-day context (2026-08-06 to 2026-08-12): across 7 recorded scans, there were 1,837 new listings and 496 regressions. New listings led on 5 scans; regressions led on 2.

Archived commentary and representative evidence

Latest scan: 383 new availability signals and 196 regressions across 59 regions.

Azure AI models: models/versions delisted (likely deprecation) (91 signals; 83 deprecation candidates, 8 recurring disappearances). SRE impact: review model deployment targets and fallback model/version choices. Coverage now ranges from 0 to 0 of 64 monitored regions. The noisiest affected signal was unavailable 98.9% of prior observations. Examples: australiaeast · cohere.cohere-command-r-08-2024.1; australiaeast · cohere.cohere-command-r-plus-08-2024.1; australiaeast · microsoft.mai-thinking-1.2026-06-01; brazilsouth · cohere.cohere-command-r-08-2024.1; and 87 more.

VM SKUs: VM sizes withdrawn (105 signals; 9 deprecation candidates, 96 recurring disappearances, up to 1 prior disappearance). SRE impact: capacity planning and SKU fallback lists should be rechecked. Coverage now ranges from 12 to 42 of 64 monitored regions. The noisiest affected signal was unavailable 84.6% of prior observations. Examples: jioindiawest · m208ms.v2; jioindiawest · m208s.v2; jioindiawest · m416.208ms.v2; jioindiawest · m416.208s.v2; and 101 more.

AKS extensions: extension types now listed (59 signals; 59 net-new regional availabilities). SRE impact: new managed cluster capabilities such as GitOps, policy, or observability. Coverage now ranges from 59 to 59 of 64 monitored regions. The noisiest affected signal was unavailable 100.0% of prior observations. Expansion pattern: first observed anywhere in the monitored regions. Examples: australiacentral · bluechiptech.bdp; australiacentral2 · bluechiptech.bdp; australiaeast · bluechiptech.bdp; australiasoutheast · bluechiptech.bdp; and 55 more.

Azure AI models: newer models/versions rolling out (28 signals; 28 net-new regional availabilities). SRE impact: new regional model/version options for latency, residency, and model selection. Coverage now ranges from 31 to 31 of 64 monitored regions. The noisiest affected signal was unavailable 100.0% of prior observations. Expansion pattern: first observed in Africa; first observed in Asia; first observed in Middle East; first observed in Oceania; first observed in South America; regional expansion of an existing signal. Examples: australiaeast · black-forest-labs.flux-2-flex.1; brazilsouth · black-forest-labs.flux-2-flex.1; canadacentral · black-forest-labs.flux-2-flex.1; canadaeast · black-forest-labs.flux-2-flex.1; and 24 more.

VM SKUs: VM sizes now offered (296 signals; 296 net-new regional availabilities). SRE impact: more compute shapes for right-sizing, performance, and cost tuning. Coverage now ranges from 14 to 22 of 64 monitored regions. The noisiest affected signal was unavailable 100.0% of prior observations. Expansion pattern: first observed in Oceania; regional expansion of an existing signal. Examples: australiaeast · d128nds.v6; australiaeast · d128nlds.v6; australiaeast · d128nls.v6; australiaeast · d128ns.v6; and 292 more.

Engineering context

  • australiaeast aiModels.cohere.cohere-command-r-08-2024.1 available → None deprecation candidate
    unavailable 0/91 prior days (0.0%)coverage 0/64 monitored regions (0.0%, -31 regions)coverage removed from 100.0% of prior regions
    Use this to evaluate model capabilities, regional deployment options, latency, and data residency. Azure OpenAI model availability
  • australiaeast aiModels.cohere.cohere-command-r-plus-08-2024.1 available → None deprecation candidate
    unavailable 0/91 prior days (0.0%)coverage 0/64 monitored regions (0.0%, -31 regions)coverage removed from 100.0% of prior regions
    Use this to evaluate model capabilities, regional deployment options, latency, and data residency. Azure OpenAI model availability
  • australiaeast aiModels.microsoft.mai-thinking-1.2026-06-01 available → None deprecation candidate
    unavailable 90/91 prior days (98.9%)coverage 0/64 monitored regions (0.0%, -29 regions)coverage removed from 100.0% of prior regions
    Use this to evaluate model capabilities, regional deployment options, latency, and data residency. Azure OpenAI model availability
  • brazilsouth aiModels.cohere.cohere-command-r-08-2024.1 available → None deprecation candidate
    unavailable 0/91 prior days (0.0%)coverage 0/64 monitored regions (0.0%, -31 regions)coverage removed from 100.0% of prior regions
    Use this to evaluate model capabilities, regional deployment options, latency, and data residency. Azure OpenAI model availability
  • brazilsouth aiModels.cohere.cohere-command-r-plus-08-2024.1 available → None deprecation candidate
    unavailable 0/91 prior days (0.0%)coverage 0/64 monitored regions (0.0%, -31 regions)coverage removed from 100.0% of prior regions
    Use this to evaluate model capabilities, regional deployment options, latency, and data residency. Azure OpenAI model availability
  • brazilsouth aiModels.microsoft.mai-thinking-1.2026-06-01 available → None deprecation candidate
    unavailable 90/91 prior days (98.9%)coverage 0/64 monitored regions (0.0%, -29 regions)coverage removed from 100.0% of prior regions
    Use this to evaluate model capabilities, regional deployment options, latency, and data residency. Azure OpenAI model availability
  • canadacentral aiModels.cohere.cohere-command-r-08-2024.1 available → None deprecation candidate
    unavailable 0/91 prior days (0.0%)coverage 0/64 monitored regions (0.0%, -31 regions)coverage removed from 100.0% of prior regions
    Use this to evaluate model capabilities, regional deployment options, latency, and data residency. Azure OpenAI model availability
  • canadacentral aiModels.cohere.cohere-command-r-plus-08-2024.1 available → None deprecation candidate
    unavailable 0/91 prior days (0.0%)coverage 0/64 monitored regions (0.0%, -31 regions)coverage removed from 100.0% of prior regions
    Use this to evaluate model capabilities, regional deployment options, latency, and data residency. Azure OpenAI model availability
  • canadacentral aiModels.microsoft.mai-thinking-1.2026-06-01 available → None deprecation candidate
    unavailable 90/91 prior days (98.9%)coverage 0/64 monitored regions (0.0%, -29 regions)coverage removed from 100.0% of prior regions
    Use this to evaluate model capabilities, regional deployment options, latency, and data residency. Azure OpenAI model availability
  • canadaeast aiModels.cohere.cohere-command-r-08-2024.1 available → None recurring disappearance
    unavailable 0/91 prior days (0.0%)coverage 0/64 monitored regions (0.0%, -31 regions)coverage removed from 100.0% of prior regions
    Use this to evaluate model capabilities, regional deployment options, latency, and data residency. Azure OpenAI model availability

Generated from the monitor's structured change evidence. See the status meanings for what each signal proves, or browse all posts.

Help make this page clearer

Two short answers; GitHub will open a draft for your review. No Forms, token, or automatic submission.

This repository is public. Review the text and image for private details. Opening the draft sends text/context to GitHub in its URL. Pasting or attaching an image uploads it to GitHub immediately.

Describe the desired outcome in plain text. This becomes the issue's ### Objective. After reviewing the issue, a maintainer can add azure-backlog to queue it for a scheduled agent; the website does not add that label or start a run.

Page context included in the draft

Nothing is posted by this website. Sign in to GitHub if needed, review, attach your screenshot, and choose Submit. Closing this panel clears the local draft and screenshot.