This optional research check is not the main feedback path.
For maintainer observations, use page feedback instead.
Founder or repeat-reader observations are qualitative, not unbiased first-time-reader results.
Read the briefing, then answer from memory: what changed most, where, and what does the evidence prove?
This measures comprehension, not whether you liked the design.
Privacy: answers and timing stay in this page's memory.
No analytics, cookies, background submissions, or browser storage.
Ordinary page/evidence requests still reach the static host.
Opening a GitHub draft sends its summary to GitHub, but does not submit an issue.
You review it and choose the final Submit on GitHub.
Public feedback: submitted issues and attachments in this repository are public.
Do not share secrets or private information. The dashboard needs no backend or authentication token.
Use one attempt. Switching away from this tab marks it interrupted; it does not silently pause the clock.
You may stop or skip at any time. Do not enter names, email addresses, subscription IDs, or private workload details.
No screenshot capture is offered in this timed check. Use normal page feedback for optional screenshots.
Case: 2026-10-03. The timer limits reading, not answering.
This is a self-selected check, not a controlled experiment.
Choose your perspective and prior familiarity to begin.
15 seconds
Since the previous scan Scan-wide briefing
Evidence gaps need attention before interpreting regional changes
2026-10-02 to 2026-10-03
39 regions with new changes or gaps
Australia East, Brazil South, Canada Central, Canada East, Central India, Central US, East Asia, East US, East US 2, East US 2 EUAP, France Central, Germany West Central, India South Central, Italy North, Japan East, Japan West, Korea Central, North Central US, North Europe, Norway East, Norway West, Poland Central, South Africa North, South Central US, South East Asia, South India, Spain Central, Sweden Central, Switzerland North, Switzerland West, UAE North, UK South, UK West, West Central US, West Europe, West US, West US 2, West US 3, github-global
163New listings
296New delistings
218Known restorations
1,449Tracked absences
13Evidence gaps
Absence tracking starts at 2026-09-05; older disappearances may not be included.
What this proves: read-only catalog/list or measurement evidence.
Not quota, capacity, successful deployment, an outage, or confirmed retirement.
Methodology
Changes relevant to you
One listing = one feature in one region
All regions and services. Groups below separate new changes from continuing observations.
A GitHub Models endpoint inference-latency observation, not an Azure regional model-catalog result. A model identifier observed by this modality; exact capabilities are not verified.
This records an inference request to the configured GitHub Models endpoint.
Context limits, input/output modalities, licensing, and tool support need exact-model review.
Timing describes the configured request and runner-to-endpoint path, not universal model speed or an Azure region. GitHub documents GitHub Models as retired on July 30, 2026; retained observations do not prove the service is currently available. No exact model capability or version is verified. The source provides category context only. Regional model-catalog evidence does not establish quota, account approval, deployment type, inference access, content filtering, performance, or data residency.
No trustworthy current result. Do not treat this as an unavailable service.
VM SKUs
1,419VM sizes
1,611 feature-region records
New listings
160 feature-region records
South Africa North
Example: Standard D128ads V7 VM size. Listing coverage: 27 → 29 regions.
New regional listing of an already observed feature.
Exact capability unverified
Standard D128ads V7 VM size
VM catalog entry with unverified exact specifications.
This signal represents a name in the regional VM size catalog, not a running VM.
Exact CPU, memory, disk, and network specifications are not verified for this identifier.
Exact SKU specifications have not been verified. A regional SKU listing does not establish quota, capacity, price, successful deployment, or measured performance. Temporary disks are not durable application storage.
Catalog/list evidence only; verify exact requirements before changing regional plans.
Known restorations
218 feature-region records
Canada East, East US, West Europe
Example: Standard D128ads V7 VM size. Listing coverage: 27 → 29 regions.
Exact capability unverified
Standard D128ads V7 VM size
VM catalog entry with unverified exact specifications.
This signal represents a name in the regional VM size catalog, not a running VM.
Exact CPU, memory, disk, and network specifications are not verified for this identifier.
Exact SKU specifications have not been verified. A regional SKU listing does not establish quota, capacity, price, successful deployment, or measured performance. Temporary disks are not durable application storage.
Catalog/list evidence only; verify exact requirements before changing regional plans.
New delistings
295 feature-region records
India South Central, West US 3
Example: Standard M128dms V2 VM size. Listing coverage: 53 → 52 regions.
Exact capability unverified
Standard M128dms V2 VM size
VM catalog entry with unverified exact specifications.
This signal represents a name in the regional VM size catalog, not a running VM.
Exact CPU, memory, disk, and network specifications are not verified for this identifier.
Exact SKU specifications have not been verified. A regional SKU listing does not establish quota, capacity, price, successful deployment, or measured performance. Temporary disks are not durable application storage.
If this matches a planned regional target, verify it with the provider. This is not an outage or retirement notice.
Tracked absences
938 feature-region records
Brazil South, Central India, East US 2, East US 2 EUAP, France Central, India South Central, Italy North, Japan East, North Europe, Norway East, Norway West, South Africa North, South Central US, South East Asia, Switzerland North, UK South, West Europe, West US, West US 2
Example: Standard Nv12s V2 VM size. Listing coverage: 6 → 6 regions.
Exact capability unverified
Standard Nv12s V2 VM size
VM catalog entry with unverified exact specifications.
This signal represents a name in the regional VM size catalog, not a running VM.
Exact CPU, memory, disk, and network specifications are not verified for this identifier.
Exact SKU specifications have not been verified. A regional SKU listing does not establish quota, capacity, price, successful deployment, or measured performance. Temporary disks are not durable application storage.
Previously tracked delistings remain absent. No new delistings does not mean recovery.
Azure AI models
20AI models
515 feature-region records
New listings
3 feature-region records
Japan West, UK West, West US 2
Example: GPT-oss 20b model from OpenAI (version 11). Listing coverage: 28 → 31 regions.
New regional listing of an already observed feature.
Category context
GPT-oss 20b.11 - OpenAI
A model identifier observed by this modality; exact capabilities are not verified.
Catalog presence identifies a provider/model/version candidate, not its tested abilities.
Context limits, input/output modalities, licensing, and tool support need exact-model review.
No exact model capability or version is verified. The source provides category context only. Regional model-catalog evidence does not establish quota, account approval, deployment type, inference access, content filtering, performance, or data residency.
Catalog/list evidence only; verify exact requirements before changing regional plans.
New delistings
1 feature-region record
Central India
Example: GPT-5 1 codex model from OpenAI (version 2025-11-13). Listing coverage: 30 → 29 regions.
Category context
GPT-5 1 codex.2025 11 13 - OpenAI
A model identifier observed by this modality; exact capabilities are not verified.
Catalog presence identifies a provider/model/version candidate, not its tested abilities.
Context limits, input/output modalities, licensing, and tool support need exact-model review.
No exact model capability or version is verified. The source provides category context only. Regional model-catalog evidence does not establish quota, account approval, deployment type, inference access, content filtering, performance, or data residency.
If this matches a planned regional target, verify it with the provider. This is not an outage or retirement notice.
Tracked absences
511 feature-region records
Australia East, Brazil South, Canada Central, Canada East, Central US, East Asia, East US, East US 2, France Central, Germany West Central, Italy North, Japan East, Japan West, Korea Central, North Central US, North Europe, Norway East, Poland Central, South Africa North, South Central US, South East Asia, South India, Spain Central, Sweden Central, Switzerland North, Switzerland West, UAE North, UK South, UK West, West Central US, West Europe, West US, West US 2, West US 3
Example: llama 3 2 11b vision instruct model from Meta (version 1). Listing coverage: 0 → 0 regions.
Category context
llama 3 2 11b vision instruct.1 - meta
A model identifier observed by this modality; exact capabilities are not verified.
Catalog presence identifies a provider/model/version candidate, not its tested abilities.
Context limits, input/output modalities, licensing, and tool support need exact-model review.
No exact model capability or version is verified. The source provides category context only. Regional model-catalog evidence does not establish quota, account approval, deployment type, inference access, content filtering, performance, or data residency.
Previously tracked delistings remain absent. No new delistings does not mean recovery.
No matching change or evidence-gap records. This is not a deployment-health assessment.
Observation history and comparison coverage
1,449 tracked listings remain absent. Zero new delistings does not mean previous absences recovered. Absence tracking starts at 2026-09-05; older disappearances may not be included.
Previous snapshot: 2026-10-02T10:03:27.302284+00:00 Current snapshot: 2026-10-03T09:23:19.985772+00:00
Snapshot timestamps do not establish freshness of every carried-forward modality. The monitor does not assess your running workloads.
Added check records: 0; removed check records: 0.
Changes to the observation set do not establish provider intent.
Exact changes and supporting evidence
Grouped by feature. Coverage counts are scan-wide; region and service filters limit the listed records.
Open to load the complete records, without rerunning probes.
This is not submitted. Open a GitHub draft to review a human-readable summary
of your answers, case/view IDs, timing, and eligibility. No raw JSON packet goes in the URL.
Only your final Submit on GitHub creates the public issue.
The full local record includes your answers, timing, role, case ID, and presentation version,
with no expected answers or automatic grade. You may download JSON and manually attach it
to the draft. Review it first: attachments upload immediately when pasted or attached on GitHub,
even before you submit the issue.
GitHub may log the draft URL and requires its own sign-in to submit. The dashboard never asks for a token.
If the summary exceeds the draft URL limit, the full local result and JSON download remain available;
nothing is truncated or automatically sent.