Rubric
Contents — domains, guide and mocks

The Claude model family

CCAO-F 3.213 min read · checked 21 September 2026

Task statementDifferentiate between Claude model types (Haiku, Sonnet, Opus)

The tiers, and what moves as you go down

faster and cheaper at the top · more capable and costlier below

  1. Haikufastest, cheapest, near-frontier quality
  2. Sonnetthe balance of speed and intelligence
  3. Opuslong-running, complex, agentic work
Capability, cost and latency all rise together. That is the whole trade-off, and it is why the most capable model is not automatically the right one.

How the exam frames it

The task statement names three model types, and that three-way split is the mental model to answer with: Haiku is the speed-and-cost tier, Sonnet is the everyday balance, Opus is the capability tier. Almost every item in this objective is a restatement of that. If a scenario stresses volume, latency or budget, it is pointing at Haiku; if it stresses hard multi-step reasoning, long autonomous work or accuracy on something difficult, it is pointing at Opus; if it describes ordinary professional work, it is pointing at Sonnet.

The documented positioning of the current models, in Anthropic’s own words, is worth reading once. Claude Fable 5.1 is for demanding reasoning and long-horizon agentic work. Claude Opus 5.5 is for long-running agentic coding and knowledge work. Claude Sonnet 5 is described as the best combination of speed and intelligence. Claude Haiku 4.5 is the fastest model, with near-frontier intelligence — which is the phrase to remember, because it contradicts the common assumption that the cheap tier is the weak one.

ModelTierContext windowPrice per million tokens
Claude Fable 5.1Fable1M$10 in · $50 out
Claude Opus 5.5Opus1M$4 in · $20 out
Claude Sonnet 5Sonnet1M$2 in · $10 out
Claude Haiku 4.5Haiku200K$1 in · $5 out

Two things in that table repay attention. The spread from the cheapest to the most expensive is an order of magnitude, which is why tier choice is a real budgeting decision for anyone running work at volume rather than a preference. And the context window is not uniform: the larger models document a one-million-token window while Haiku documents two hundred thousand. Token prices matter mostly to people building on the API; in the Claude apps you are spending a usage allowance rather than a per-token bill, but the same ordering holds — heavier models consume it faster.

What actually differs between tiers

It helps to separate four properties that people tend to bundle into the single word “better”.

PropertyWhat it meansHow it moves across the tiers
CapabilityDepth on hard, multi-step or ambiguous problemsRises from Haiku to Sonnet to Opus and above
SpeedHow quickly the answer comes backHaiku is documented as the fastest
CostTokens on the API; usage allowance in the appsRoughly ten times from the cheapest to the dearest
Context windowHow much can be in front of the model at once200K on Haiku; 1M on the larger current models

Note what is not on that list. Tiers do not differ in what kind of work they will attempt, in tone, in the languages they handle, or in their willingness to follow instructions. All the current models take text and image input, produce text, work across languages and use tools. Choosing Opus does not unlock a feature; it buys depth on problems where depth is what was missing.

What “better” is actually buying

Changes with the tier

  • Depth on hard reasoning
  • Speed of response
  • Cost per unit of work
  • Size of the context window

Does not change with the tier

  • Whether it knows your company’s figures
  • Whether a claim has been verified
  • Whether the brief was clear
  • Whether it can see a document you did not attach
The left column is what changes with the tier. The right column does not — so a model choice never fixes a problem in that column.

Choosing in the Claude apps

In the apps the model is chosen from the selector next to the send button — click the model name to switch, and “More models” to see additional options. Alongside it sit two settings that change behaviour within a model, and confusing them with the tier is a common mistake.

  • Effort determines how thorough a response is and how much of your usage it consumes. Low and medium suit routine tasks and stretch your usage further; high is the default and is described as the best overall balance of quality and speed; extra high and max exist for long-running and deeply demanding work.
  • Thinking is a separate toggle for extended reasoning, showing an expandable section above the answer. On the most capable current models it cannot be turned off at all.
  • Administrators on Enterprise plans can restrict which models are available, so the picker is not the same everywhere.

The help centre’s advice on these is refreshingly plain: simple questions, basic information requests and general writing do not need extra effort or thinking, while mathematical problems, coding challenges, project planning and technical analysis are where raising effort or turning thinking on is worth it. In other words, the same discipline as tier choice, one level down.

This objective is about telling the tiers apart. Matching a specific task’s requirements — cost, speed and quality together, with the trade-offs that involves — is objective 3.3, and the context-window consequences of a very long conversation are 3.4.

Traps the wrong answers are built from

Tempting but wrongDo this instead
Treating the most capable model as the default for everythingMatch the tier to what the task needs; volume and latency often point the other way.
Assuming the fast tier is the low-quality tierHaiku is documented as fastest with near-frontier intelligence, not as a weak model.
Switching models to fix a vague prompt or a missing fileCheck the source and the brief first; a tier change fixes neither.
Confusing the effort setting with the model tierEffort changes thoroughness within a model; the tier changes the model.
Assuming every user sees the same model listAvailability varies by plan, and Enterprise administrators can restrict it.

You should now be able to

  • State the defining property of each tier: Haiku speed and cost, Sonnet balance, Opus capability.
  • Recognise which tier a scenario is describing from its constraints rather than its adjectives.
  • List the properties that change across tiers — capability, speed, cost, context window — and those that do not.
  • Distinguish the model tier from the effort and thinking settings inside the apps.
  • Explain why a more capable model does not fix a missing source or an unclear brief.
  • Account for the gap between the exam’s three-tier framing and the current product lineup.

Practice questions

Original questions written for this lesson, in the exam’s style. Answer first, then open the reasoning — every option is explained, including why the wrong ones are tempting.

  1. Question 1

    A logistics company runs an automated step that classifies 60,000 short delivery-exception notes a day into twelve categories. The classification rules are simple, accuracy on a sample is already above 98%, and the step must complete within seconds.

    Which tier best fits this workload?

    1. AThe Opus tier, because accuracy matters in a customer-facing process.
    2. BThe Haiku tier, because volume, latency and cost are the binding constraints.
    3. CThe Sonnet tier, because it balances speed and intelligence for all workloads.
    4. DAlternating tiers so that no single model carries the whole load.
    Show answer and reasoning
    1. AIncorrect. Accuracy is already high and the task is simple, so the extra capability buys nothing while multiplying cost and latency.
    2. BCorrect. This is the documented use case for the fastest, cheapest tier: high-volume, latency-sensitive, cost-sensitive processing.
    3. CIncorrect. It is the sensible default for varied work, but here volume and latency dominate and the task is easy.
    4. DIncorrect. Alternating adds inconsistency and complexity without addressing any stated constraint.
  2. Question 2

    A colleague argues that the team should always use the most capable model available, because “if it is better at hard things it is better at easy things too, so we will never be caught out”.

    Which response is most accurate?

    1. AHe is right; the only reason to use a smaller model is when the larger one is unavailable.
    2. BHe is right about quality but wrong about tone, which differs between tiers.
    3. CCapability rises with the tier, but so do cost and latency, and easy tasks gain little from it.
    4. DSmaller models cannot handle business documents, so his rule is safe for office work.
    Show answer and reasoning
    1. AIncorrect. This ignores the cost, speed and usage-allowance differences that are the whole point of having tiers.
    2. BIncorrect. Tone is not a documented difference between tiers; all current models are general-purpose across languages and tasks.
    3. CCorrect. The trade-off is three-sided; on simple, high-volume work the lighter tier is already accurate and is faster and cheaper.
    4. DIncorrect. All current models handle text and images and are used for business documents; the difference is depth, not permission.
  3. Question 3

    A compliance team must analyse three years of incident reports against a new regulatory framework, reconciling contradictory records and producing findings that the regulator may examine. The work runs once a quarter and takes several hours of model time.

    Which TWO considerations most support choosing the capability tier here? (Select 2.)

    1. AThe reasoning is multi-step and ambiguous, which is where depth of capability actually pays.
    2. BThe task runs four times a year, so the cost difference is small against the stakes.
    3. COnly the capability tier can read long documents at all.
    4. DThe capability tier removes the need to verify the findings before submission.
    5. EUsing the top tier will make the analysis faster to produce.
    6. FThe regulator requires the most capable model available to be used.
    Show answer and reasoning
    1. ACorrect. Hard, multi-step, ambiguous analysis is the documented positioning of the top tiers.
    2. BCorrect. Cost pressure comes from volume; at four runs a year the premium is negligible next to a regulatory error.
    3. CIncorrect. All tiers read documents; context window sizes differ, but capability is not a prerequisite for reading.
    4. DIncorrect. No model tier removes verification, and regulated output is exactly where human review matters most.
    5. EIncorrect. More capable models are generally slower, not faster; speed is the lighter tiers’ advantage.
    6. FIncorrect. No such requirement is described anywhere; this is an invented constraint.
  4. Question 4

    A new user on a Team plan cannot find one of the models a colleague at another company mentioned, and assumes something is broken with her account.

    What is the most likely explanation?

    1. AModels appear only after a certain amount of usage on an account.
    2. BModel availability varies by plan, and administrators can restrict which models an organisation sees.
    3. CThe model list is identical everywhere, so her colleague is mistaken about the name.
    4. DShe needs to raise the effort setting before additional models appear.
    Show answer and reasoning
    1. AIncorrect. Nothing in the documentation ties model availability to accumulated usage.
    2. BCorrect. Enterprise plans may have restricted availability based on administrator configuration, and the selector differs by plan.
    3. CIncorrect. The list is not identical; availability differs by plan and organisation settings.
    4. DIncorrect. Effort changes thoroughness within a model and does not reveal other models.

Sources

Drafted with AI assistance and checked against the sources above; expert review is in progress. Spotted an error? Tell us and it gets fixed, dated and listed on how this is written.