LLM Pulse vs CitedSpy (2026): Honest Budget Tracker or Cheapest Unreviewed One?
LLM Pulse starts at €49/mo with published pricing and unlimited seats. CitedSpy starts around $16 to $19/mo with six engines and zero independent reviews.
LLM Pulse and CitedSpy are the two budget trackers people find when they refuse a $200 sales call. LLM Pulse is a three-person Barcelona shop with published tiers and almost no verified reviews. CitedSpy is cheaper still, with six engines on every plan and not one independent review anywhere. The two products appeal to the same buyer, a team that wants a tracker without a procurement cycle, but they answer the buyer with very different levels of disclosure.
This is a diligence comparison more than a feature one. LLM Pulse at least publishes a price list and names a team. CitedSpy has the best sticker-to-engine ratio in the category and a site that disagrees with itself on price. Fine as a $16 experiment. Do not make it the number you report to a client. If you need daily refresh and Claude, Promptwatch Essential is the more established option. If you also need crawler logs, use Promptwatch Professional. The diligence question is which of these you can defend in a QBR, and the answer depends on what each vendor publishes.
Snapshot comparison
| LLM Pulse | CitedSpy | Promptwatch | |
|---|---|---|---|
| Starting price | €49/mo annual (Starter, 50 prompts) | From $16/mo (pages also list $19) | Free Explore; $95/mo Essential |
| Engines at entry | 5: ChatGPT, Perplexity, Gemini, AI Mode, AI Overviews | 6 + AI Overviews, no per-engine surcharge | 10+ on paid plans |
| Claude / Copilot / Grok | Enterprise add-ons | Included | Included from $95/mo |
| Cadence | Weekly default; daily costs more, no trial | Up to 7×/week | Daily on paid plans |
| Seats | Unlimited on every plan | Self-serve | Scales with tier |
| Independent reviews | Essentially one verified G2 review | None | 4.7/5, 1,840+ customers |
| Best for | Small teams that want transparent pricing | Cheap experiments, not a system of record | Teams prepared to use Professional for crawler logs and a content loop |
The table is a diligence grid, not a feature grid. The rows that matter are the ones about reviews and pricing transparency, because those are the rows that tell you whether you can trust the numbers in the other rows.
LLM Pulse's case
Published six tiers, unlimited seats, MCP even on the 14-day trial, AI traffic analytics, ChatGPT Shopping tracking. Bootstrapped and profitable since July 2025. Softonic and Adevinta are named customers. Default refresh is weekly. Daily costs more. Headline prices assume annual billing. Starter is €49/mo annual for 50 prompts. Five engines: ChatGPT, Perplexity, Gemini, AI Mode, AI Overviews. Claude, Copilot, and Grok are Enterprise add-ons. Three people run the company. Essentially one verified G2 review. The disclosure is real, even when the disclosure is thin. A three-person team that names itself and names its customers is a different procurement risk from an anonymous site.
You pay more than CitedSpy for published pricing, unlimited seats, a named team, and features (MCP, shopping, traffic analytics) CitedSpy has not evidenced. Weekly cadence is the trade. If that is acceptable, LLM Pulse is the honest budget tracker. The trade is the whole value proposition: you give up daily refresh and get a price list you can put in a spreadsheet.
CitedSpy's case
Best sticker-to-engine ratio in the category. Six engines plus AI Overviews, no per-engine surcharge. Claude, Copilot, and Grok included. Cadence up to 7×/week. Entry from $16/mo. Pages also list $19. Top tier is listed as both $333 and $399. The internal price disagreement is the kind of thing a buyer notices on a second read, and it is the reason to confirm at checkout rather than at the pricing page.
Changelog has listed blank sentiment reports and 100% win-rates. "90% accuracy" is self-reported. Terms allow features to change or disappear. Zero independent reviews. The site says six-engine coverage is real. There is no third-party review to confirm the numbers work. A self-reported accuracy figure with no review behind it is a marketing number, not a measurement, and treating it as a measurement is the risk.
Fine as a $16 experiment. Do not make it the system of record. The experiment is worth running because six engines for $16 is cheap enough to verify by hand. The system of record is worth paying for, because a system of record has to be defensible.
Who should pick which
Pick LLM Pulse if you want a real published price list and unlimited seats, and weekly cadence is acceptable. You are buying honesty more than coverage. Claude waits for Enterprise. The buyer here values being able to quote a price in a meeting without a sales call.
Pick CitedSpy only to try six engines cheaply. Do not make it the number you report to a client. Confirm which price the checkout actually charges ($16 vs $19). The buyer here is running a cheap parallel check against a more established tool, not building a reporting stack on it.
If you need daily refresh and Claude, Promptwatch Essential covers ten-plus engines at $95/mo. If you also need crawler logs, Professional is the first brand plan with them at $245/mo. Promptwatch is rated 4.7/5 on G2 and has 1,840+ customers. The other two are trackers. Promptwatch is a stack. Neither budget tool replaces its visitor analytics or Content Agents. The stack is the difference between a tracker that answers "was I mentioned" and a platform that answers "was I mentioned, by what, and did it convert."
A team that refuses a $200 sales call can still start on Explore for free, then decide whether €49 weekly Pulse, $16 CitedSpy, or $95 Essential is the log they will keep. The free Explore tier is the one that lets a buyer compare without spending, which is the same impulse that brings them to the budget tools in the first place.
FAQ
Why would I pay more for LLM Pulse?
Published pricing, unlimited seats, a named team, and features (MCP, shopping, traffic analytics) CitedSpy has not evidenced.
Is CitedSpy's six-engine coverage real?
The site says so. There is no third-party review to confirm the numbers work.
Does either replace Promptwatch?
Not for crawler logs, visitor analytics, or Content Agents. They are trackers. Promptwatch is a stack.