Guides

Social Media MCP Costs and Limits: Build a Research Budget

Budget social media MCP research with exact search quotes, optional connection caps, saved-result reuse and per-item bulk costs. Includes worked examples.

Budget an MCP research task by the searches it needs, not by the length of the assistant's answer. Searching posts, opening a result, reading comments and refreshing later can be different actions. A useful estimate names those actions, shows their quoted credit costs and reserves room for the follow-up you may actually need.

In Superlurk, current catalog searches fall into 1-, 5- and 25-credit bands. The assistant asks for a price quote before running a search, and the quote is the spending reference. Most searches cost 1 credit, but a blanket assumption that every LinkedIn search costs 5 is wrong: several deeper LinkedIn actions cost 25. Check the selected action rather than budgeting from the platform name.

Credits are not the same thing as a subscription price

Credits measure search usage inside Superlurk. Subscription and top-up prices are listed on the pricing page, where you should check the current allowance, retention and monitor limits. Your AI client may also charge for its own service. A Superlurk credit quote does not include the assistant provider's subscription or model charges.

A search can also have more than one internal provider step, such as resolving an account before fetching its posts. Superlurk prices the customer action before it runs. Do not add an assumed lookup fee on top of a quote, or promise a discount because the assistant happens to use a different phrase.

Make a small budget before the first request

Before running anything, propose the smallest useful search plan. List each network, action, required input and quoted credits. Prefer my saved results where they answer the question. Ask me before spending, stop at 12 credits in total, and ask again before adding another result page or refresh.
Illustrative prompt, not a recorded search or customer result

That prompt is a useful instruction, but it is not an account-enforced cap. Superlurk also offers an optional daily credit cap per connection. Review it when connecting and under your connection settings later. A cap has no value if you leave it unset and then assume a sentence in a chat enforces it everywhere.

Worked example: a ten-credit customer-question sample

Imagine you are deciding which onboarding problems deserve interviews. You want examples from public posts, not a representative survey. If the current quotes for the chosen actions are all 1 credit, this plan illustrates a small first pass:

  • Two keyword searches on two selected networks: 2 credits.
  • Open or inspect two promising items: 2 credits if those actions quote 1 each.
  • Retrieve comments for two items: 2 more credits on the same assumption.
  • Keep 4 credits uncommitted for a useful follow-up or additional page.

The total planned retrieval is 6 credits with 4 in reserve. This is an illustration, not a fixed package or a claim that six calls produce a complete answer. If a chosen search quotes 5 or 25, revise the plan before approval. A smaller sample may be enough to frame interview questions; it is not enough to estimate how common a complaint is.

Bulk lookup: one quote does not mean one item price

Eligible bulk lookup accepts up to 50 accounts, links or IDs for the chosen search. Shared settings apply to the list, while every valid item is validated, priced and run like a single lookup. The quote identifies skipped duplicates and invalid inputs. A failed item is free, but successful items can still be charged when another item fails.

If 12 pasted profile links contain 2 exact duplicates and 1 invalid value, the quote may propose 9 lookups. If the selected action quotes 5 credits per lookup, the total is 45 credits. If it quotes 1, the total is 9. Inspect the actual quote: input normalization, your saved results and current in-flight work can affect which items need a new billable run.

Do not split a list into several runs to evade a cap or mistake the single approval for a flat-rate batch. For preparation and result review, use the bulk social profile lookup guide.

Know what is free and what needs a new quote

  • Reopening your own saved result is free. Ask the assistant to use it before retrieving the same evidence again.
  • A failed search is free. Read the result status; a valid empty result is not automatically a technical failure.
  • Cancellation before sending is free. Do not assume cancelling after retrieval has started reverses a charge.
  • A refresh is new retrieval. Review its quote rather than assuming an old result entitles you to unlimited updates.
  • Another page or a follow-up can be another action. A button that fills in a reply or detail search does not run it for free.

Separate a credit cap from a research stopping rule

A cap controls spend. A stopping rule controls when you have enough evidence for the decision. Without the second, an assistant can keep returning more examples without improving the work. Decide in advance whether you need five source-checked problems, coverage of three communities or a short comparison across two networks.

Require the assistant to say when coverage is incomplete. If a source is unavailable, spend the remaining budget only when a different source would answer the same question. Do not quietly substitute a web search for a direct comment lookup and present the resulting sample as equivalent.

Keep a record that makes the bill understandable

Save the research question, planned total, approved quote and result statuses with the final brief. Note which items were reused and which were retrieved again. Keep your own assistant subscription charges separate so a later reviewer does not confuse model usage with Superlurk credits.

For a wider introduction, read how social media MCP research works. For the plans and current cash prices, use Superlurk pricing. The best budget is the one you can explain before approving it and reconcile after the task finishes.

Frequently asked questions

Are all LinkedIn lookups 5 credits?

No. Current LinkedIn actions have different prices, including 5- and 25-credit searches. The selected action's quote shows the exact cost before running.

Does one bulk lookup approval cost one credit?

No. A bulk quote totals the billable lookups it contains. Every valid item is priced individually; skipped duplicates and invalid items are identified separately.

Does a prompt budget replace the connection cap?

No. A prompt is an instruction to the assistant. An optional connection-level daily cap is an account control. Review both rather than assuming either is automatically set.

Sources

  1. Model Context Protocol: tool interaction model

How this guide was written. Written for Superlurk's public-data search application on 8 November 2026, using its current catalog, MCP contract and the primary references below. Prompts, budgets and sample tables are illustrative; they are not measured customer outcomes. Search availability and quoted prices can change.

The Superlurk TeamSocial research

We build Superlurk for public-data research across 25 supported social networks, with the price shown before every search. We write practical guides to search methods, source verification and AI-assisted research.

Start lurking for free.

Every network, no card needed. See the price first, and pay nothing if a search fails.