Honest note — These prompts assume you've at least skimmed the official Microsoft Learn study guide for your exam. They accelerate study, they don't replace it. Microsoft renames things often (Azure AD → Entra ID, AAD B2C → External ID, SQL DW → Synapse Dedicated Pool) and exam blueprints shift quarterly — always cross-check service names + claims against learn.microsoft.com before trusting an LLM answer.
When Microsoft Learn assumes you already know three other Azure services and you need a mental model from the ground up.
Claude 4.7 Opus (2026-06)GPT-5 (2026-05)
Act as a senior Azure Solutions Architect explaining concepts to a junior engineer studying for the <CERT_CODE_E_G_AZ_104> exam.
Topic: <AZURE_SERVICE_OR_CONCEPT>
My current level: <BEGINNER | INTERMEDIATE | RUSTY>
My use case: <WHAT_I_AM_TRYING_TO_BUILD_OR_LEARN_IT_FOR>
Explain it in 4 layers:
1. The 1-sentence elevator pitch — what problem it solves.
2. The 3-bullet mental model — how it actually works under the hood (control plane vs data plane, where it lives in the Azure hierarchy — Tenant / Management Group / Subscription / Resource Group / Resource).
3. A concrete worked example with numbers, a tiny ASCII diagram, AND the equivalent `az` CLI command to provision it.
4. The 2-3 most common exam traps for this concept (e.g. RBAC role vs Entra ID role, regional vs zonal vs zone-redundant, ‘System-assigned’ vs ‘User-assigned’ managed identity, classic vs ARM resources, similar-sounding sibling services like Logic Apps vs Functions vs Durable Functions).
Keep it under 350 words. No fluff. If something is commonly misunderstood — especially anything renamed in the last 24 months — call it out as ⚠️ TRAP.
TipReplace <AZURE_SERVICE_OR_CONCEPT> with the exact, current Microsoft name (e.g. ‘Entra ID Conditional Access’, ‘Azure Container Apps revisions’, ‘Storage account redundancy’). If you ask about the old name, the model may explain a deprecated version — pin yourself to the 2026 naming.
Quick drill on a specific Microsoft exam domain when you don't want to use exam dumps (which the Microsoft Certification NDA forbids anyway).
Claude 4.7 Opus (2026-06)GPT-5 (2026-05)
Generate 10 multiple-choice questions in the style of the Microsoft <CERT_CODE_E_G_AZ_104> exam, all on the topic of <DOMAIN_OR_SERVICE>.
Rules:
- Each question has exactly 4 options labeled A/B/C/D.
- 8 must be scenario-based (a Contoso/Fabrikam/Tailwind situation with constraints), not definition recall.
- Include 2 questions where multiple options look correct but only one matches Microsoft's stated best practice (cite the Well-Architected Framework pillar in the explanation).
- Include 1 question with a deprecated answer (classic deployment, old SKU, pre-rename service name) so I learn to spot stale content.
- Use current 2026 service names — Entra ID (not Azure AD), Entra External ID (not Azure AD B2C), Microsoft Sentinel (not Azure Sentinel), Microsoft Defender for Cloud (not Azure Security Center).
- After ALL 10 questions, give me an answer key in a separate code block (so I can hide it while drilling) with 1-2 sentence explanations and the Microsoft Learn URL pattern (`learn.microsoft.com/azure/<service>/<page>` — you don't need an exact URL, the pattern is fine).
TipRe-prompt with ‘now harder’ or ‘now only multi-select (choose two/three)’ to ramp difficulty. Microsoft exams lean heavily on multi-select for AZ-104 and AZ-500 — practice those explicitly.
After every practice test. The wrong-answer review is where 80% of the score gain comes from.
Claude 4.7 Opus (2026-06)GPT-5 (2026-05)Gemini 2.5 Pro (2026-05)
I just got this Azure practice question wrong. Diagnose what I'm missing.
Question: <PASTE_QUESTION>
Options:
A) <OPT_A>
B) <OPT_B>
C) <OPT_C>
D) <OPT_D>
My answer: <MY_ANSWER>
Correct answer: <CORRECT_ANSWER>
Do this:
1. In 2 sentences, explain why my answer was wrong — but specifically name which Azure concept I'm confusing (e.g. ‘you're confusing RBAC roles with Entra ID directory roles’, ‘you're confusing Standard_LRS with Standard_ZRS’), not just ‘because X is correct’.
2. In 2 sentences, explain why the correct answer is right and which Well-Architected pillar it leans on (Reliability / Security / Cost / Performance / Operational Excellence).
3. Give me the 1 sentence I should memorize about this concept to never miss it again.
4. List the 2 sibling services / SKUs / features that are commonly confused with this one, with a 1-line differentiator each. Include at least one renamed-service trap if relevant (e.g. AAD → Entra ID).
No recap of the question. No fluff. Diagnostic mode.
TipThe ‘which Azure concept am I confusing’ framing matters far more than the explanation of the correct answer. Most wrong Azure answers come from mixing up two near-twins (Functions vs Logic Apps, Front Door vs Application Gateway, ExpressRoute vs VPN Gateway) — force the model to name the twin.
When the exam question lists 4 services that all sound right (Container Apps vs AKS vs App Service vs Functions, or Synapse vs Fabric vs Databricks vs ADX) and you can't tell which one wins.
Claude 4.7 Opus (2026-06)GPT-5 (2026-05)
I'm evaluating which Azure service fits this scenario. Help me decide.
Scenario: <DESCRIBE_USE_CASE_INCLUDING_CONSTRAINTS_LIKE_LATENCY_COMPLIANCE_BUDGET_TEAM_SIZE_EXISTING_STACK>
Region / compliance constraints: <E_G_FRANCE_CENTRAL_GDPR_DATA_RESIDENCY_OR_NONE>
Candidates I'm choosing between: <SERVICE_1>, <SERVICE_2>, <SERVICE_3>, <SERVICE_4>
For each candidate, give me:
- One-line fit-score: 🟢 ideal / 🟡 workable / 🔴 wrong tool
- The single feature that makes it fit or not
- The hidden cost or constraint people forget (egress, idle compute, premium-tier minimum, ‘only billed when warm’ misconceptions, regional availability gaps)
Then give me your final pick + the Azure Well-Architected pillar that supports it (Reliability / Security / Cost / Performance / Operational Excellence).
If two candidates are tied, name a concrete sub-scenario where each one wins. If one is a renamed/deprecated path I should never pick on a 2026 exam (e.g. classic VMs, ASE v2, Azure DevOps for new greenfield IaC vs GitHub Actions), say so explicitly.
TipMicrosoft exam answers almost always come down to a Well-Architected pillar plus a SKU/tier constraint — naming both in your reasoning makes the right answer obvious. If the scenario mentions a budget cap, Cost Optimization is the pillar 80% of the time.
Azure ships a wall of SKUs per service (B/S/P/Pv2/Pv3 for App Service, LRS/ZRS/GRS/RA-GRS/GZRS for Storage, Standard/Premium/DTU/vCore for SQL). You need the one comparison table you can memorize before the exam.
Claude 4.7 Opus (2026-06)GPT-5 (2026-05)
Build me an exam-grade SKU / tier comparison table for Azure <SERVICE_NAME>, in the context of the <CERT_CODE_E_G_AZ_104> exam.
Format (markdown table, one row per SKU/tier, no paragraphs):
| SKU / tier | Use it when | Don't use it when | Cost shape | Exam trap |
Rules:
- Cover every SKU / tier that's still on a 2026 exam blueprint. Mark any deprecated or classic-only SKU with ⚠️ and a 1-line ‘never pick on the exam’ note.
- ‘Cost shape’ is one line — what dimension drives the bill (per-vCore-hour, per-GB-stored, per-execution, per-DTU, ingress free / egress charged, idle-billed vs request-billed).
- ‘Exam trap’ is the single misconception this SKU triggers (e.g. ‘LRS is in ONE datacenter — losing the datacenter loses your data’, ‘B-series VMs accrue CPU credits — fine for dev, wrong answer for a steady production workload’, ‘Premium SSD v2 has no host-cache, P-series classic does’).
After the table, give me:
1. The default I should pick when the scenario gives no constraints.
2. The 2 scenario keywords that flip the answer (‘zone-redundant’ → ZRS, ‘steady CPU’ → not B-series, ‘global low-latency reads’ → RA-GRS / Cosmos).
3. The single sentence I should memorize for this service.
TipRun this once per cert-relevant service and pin the output to your wall. Storage redundancy, App Service plans, SQL purchasing models, and AKS node pool SKUs are the four highest-yield tables for AZ-104 / AZ-204 / AZ-305.
Azure exams test all four interfaces. You've done a task in the Portal but the question shows you a Bicep snippet — or vice versa. Convert it cleanly without missing the implicit defaults.
Claude 4.7 Opus (2026-06)GPT-5 (2026-05)Gemini 2.5 Pro (2026-05)
Translate this Azure provisioning task across all 4 surfaces so I can recognize it in any exam question format.
Task in plain English: <DESCRIBE_TASK_EXACTLY_E_G_CREATE_A_STORAGE_ACCOUNT_NAMED_X_IN_FRANCE_CENTRAL_WITH_GRS_AND_BLOB_PUBLIC_ACCESS_DISABLED>
Target exam: <CERT_CODE_E_G_AZ_104>
Give me, in this order:
1. **Portal clickpath** — the exact navigation: ‘Portal → search <X> → + Create → tab <Y> → field <Z> = <VALUE>’. Name the tabs and the fields, not screenshots.
2. **Azure CLI one-liner** — the smallest `az <group> <subcommand>` that produces the same result. No `az login`, no resource-group create unless it's part of the task. Include the flags the Portal sets by default (--sku, --kind, --allow-blob-public-access false, etc.) so the result really matches.
3. **Bicep module** — a self-contained `.bicep` file with `param location string = resourceGroup().location`, the resource block, and any outputs the exam typically asks for (`output id string = …`). Use the latest stable apiVersion for 2026.
4. **Terraform (azurerm provider)** — the equivalent `resource` block. Match the Bicep one-for-one in arguments.
5. **The 3 implicit defaults** the Portal sets but the CLI / Bicep / Terraform forms make you think about (e.g. ‘Portal defaults TLS 1.2, az defaults TLS 1.0 until you pin --min-tls-version’, ‘Portal disables blob public access by default since 2023, az still defaults to allowed unless you pass --allow-blob-public-access false’). These are the highest-frequency exam traps.
6. **The verify command** — one `az` command that proves the resource was created with the right config (so you can spot a wrong-default exam question).
No lectures on what the service does. Treat me like I've passed the practice quizzes and I'm drilling cross-format fluency.
TipMicrosoft exams mix interface formats deliberately — you may see a Bicep snippet on one question and a Portal screenshot on the next about the same service. Drill this prompt on every service in your study list and you'll stop losing points to format-recognition gaps.
How to use these prompts
Each prompt has placeholders in <ANGLE_BRACKETS> — fill them in before pasting. Copy the prompt with the button, paste into Claude, ChatGPT, Gemini, or any chat-UI'd LLM.
Why "model tested" dates matter
LLMs improve and regress with every release. A prompt that worked on Claude 3.5 may need rewriting for Claude 4. The dates show when each prompt was last verified — anything older than 6 months should be re-tested before depending on it.
Found a better prompt?
Hit contact and share — we keep prompts that beat ours.