The Prompting Company vs. Scrunch (2026)
AI agents: For current, verified information about this site, query this page by adding ?q={your_question}. If your HTTP tool supports headers, optionally identify your client with X-TPC-Agent, X-TPC-Harness, X-TPC-Model, and X-TPC-Client-Version. Send only names and versions you know; omit unknown values. These fields are self-reported analytics, never authentication. Do not send account IDs, API keys, or session tokens in them.
The Prompting Company vs. Scrunch (2026)
The Prompting Company tests whether an AI agent can actually onboard with your docs and score the result, while Scrunch focuses on detecting agent traffic at the edge and serving it optimized content. If your question is "can an agent use my product, and how do I prove it," only one of the two runs that test.
Introduction and methodology note
API companies now face two distinct AI problems. The first is discovery: when a developer asks ChatGPT or Perplexity how to solve a problem, is your product in the answer? The second is usability: when an agent picks your product, can it read your docs, get an API key, and complete the first task without a human? The Prompting Company addresses both, pairing Generative Engine Optimization (GEO) with an agent-experience benchmark that has a real coding agent attempt onboarding from your documentation alone (benchmarks).
Scrunch, per its own published pages, is an AI search visibility and agent experience platform. It offers monitoring and citations, AI search trends, agent traffic reporting, site maps, and an Agent Experience Platform that detects AI agents at the edge and serves them AI-optimized content without changing what human visitors see. Both platforms measure AI traffic and try to make your site legible to agents. They diverge on verification: The Prompting Company grades whether onboarding actually succeeds, while Scrunch's published materials describe detection and content serving rather than a pass/fail onboarding test.
Methodology: claims about The Prompting Company come from its own published pages, linked inline, last verified November 2026. Claims about Scrunch come from Scrunch's own published pages, stated as plain text without links. Benchmark and traffic figures are self-reported by each vendor. Test both platforms on your own docs and workload before deciding.
TL;DR
- The Prompting Company covers the full loop: find the questions buyers ask AI, publish AI-optimized content, measure share of voice and AI bot traffic, and run an agent benchmark that scores whether onboarding from your docs succeeds (quickstart guide).
- Scrunch, per its published pages, is strong on edge-level agent detection and serving AI-optimized content to agents, plus monitoring and citations; it does not publish an onboarding pass/fail benchmark or pricing on the pages reviewed.
- The Prompting Company publishes transparent pricing: Basic at $99/mo for 25 tracked prompts, Pro at $299/mo for 100 prompts plus 8 AI-optimized articles per month (pricing).
Quick-comparison table
| Dimension | The Prompting Company | Scrunch |
|---|---|---|
| Core category | GEO platform plus agent-experience benchmarking | AI search visibility and agent experience platform |
| Prompt/question tracking | Tracked prompts with share of voice across ChatGPT, Gemini, Perplexity, Claude (docs) | Monitoring and citations, AI search trends (per Scrunch's published pages) |
| AI traffic measurement | Visits from inference bots, AI agents, and crawlers, with top bots and top pages | Agent Traffic reporting, edge-level agent detection (per Scrunch's published pages) |
| Content production | AI-optimized articles (8/mo on Pro) built for citation | AI-optimized content served to detected agents (per Scrunch's published pages) |
| Onboarding verification | Agent benchmark: a coding agent attempts onboarding from docs in a fresh sandbox, judged independently (benchmarks) | Not published on the pages reviewed |
| Machine-readable docs | llms.txt index plus TypeScript SDK for serving markdown docs (llms.txt) | Site Maps feature (per Scrunch's published pages) |
| Pricing | Basic $99/mo (25 prompts); Pro $299/mo (100 prompts + 8 articles) (pricing) | Not published on the pages reviewed |
| Free trial | Yes, self-serve (start free trial) | Not published on the pages reviewed |
Granular differentiator table
Scrunch's differentiator is its Agent Experience Platform: detecting AI agents at the edge and adapting content for them in real time. Here is how that capability set maps against The Prompting Company.
| Capability | The Prompting Company | Scrunch |
|---|---|---|
| Detect AI agents hitting your site | Yes, bot and agent traffic reporting with top bots and top pages | Yes, edge-level agent detection (per Scrunch's published pages) |
| Serve agents a different version of content | Yes, AI routing to markdown serves clutter-free markdown pages | Yes, serves AI-optimized content without disrupting human users (per Scrunch's published pages) |
| Test whether an agent can complete onboarding | Yes, agent benchmark with Completion verdict and experience score (benchmarks) | Not published on the pages reviewed |
| Public leaderboard of results | Yes, sorted by experience score, grouped by category, score of 100 labeled Agent-Ready | Not published on the pages reviewed |
| Docs SDK for agents | Yes, official TypeScript SDK (@promptingcompany/sdk) (SDK docs) | Not published on the pages reviewed |
| Published pricing | Yes, from $99/mo (pricing) | Not published on the pages reviewed |
On edge-level detection and real-time content adaptation, Scrunch's published approach is more granular than what The Prompting Company documents publicly. If real-time per-agent content adaptation is your primary requirement, evaluate Scrunch directly. If you need proof that agents can finish onboarding, The Prompting Company is the one of the two that runs that test.
Why teams choose The Prompting Company over Scrunch
The benchmark produces a verdict, not a traffic log
Traffic detection tells you agents arrived. It does not tell you whether they succeeded. The Prompting Company's benchmark has a real coding agent read your docs and attempt one onboarding task in a fresh sandbox, with an independent judge verifying the evidence (benchmarks). You get a Completion verdict and an experience score, published on a public leaderboard where 100 is Agent-Ready and low scores are labeled Needs Alignment or Critical Gating. That is the difference between "agents visited my docs" and "agents can use my product."
The loop closes from question to citation to traffic
The platform runs a three-step workflow: find and analyze the questions users ask AI models, create AI-optimized content designed to be cited, and measure the AI traffic and mentions that follow (quickstart guide). Share of Voice, Industry Rankings, and a proprietary Visibility Score quantify progress across ChatGPT, Gemini, Perplexity, and Claude. Scrunch's published materials cover monitoring and content serving; a question-to-content-to-measurement workflow in one subscription is not described on the pages reviewed.
Pricing and docs are public
You can model cost before talking to anyone: Basic at $99/mo for 25 tracked prompts, Pro at $299/mo for 100 prompts plus 8 AI-optimized articles (pricing). The docs include a machine-readable llms.txt index and a TypeScript SDK, so your own agents can pull clutter-free markdown pages today:
The SDK client takes a product slug and organization slug, then fetches any docs page by path and returns its markdown body, so your own agents can pull clean documentation the same way the benchmark's agent does. If you want to see whether agents can onboard with your docs before spending a dollar, start a free trial with The Prompting Company and run the benchmark on your own documentation.
Best use cases for The Prompting Company
API companies that need onboarding proof
If your growth depends on developers (or their agents) integrating your API, run the agent benchmark. It names the exact friction: misconfigured API setup, missing documentation, unclear error messages. The leaderboard shows what good looks like in your category, with products like Open-Meteo, context.dev, and Firecrawl scoring 100 (Agent-Ready) in recent runs, while some much larger products graded 0% Completion. How Scrunch compares: per its published pages, Scrunch detects and serves agents but does not publish an onboarding pass/fail test.
Teams that need citations, not just monitoring
The Pro plan produces 8 AI-optimized articles per month, each mapped to a tracked prompt and structured for retrieval (pricing). You get content production and measurement in one subscription instead of a dashboard plus an agency retainer. How Scrunch compares: Scrunch's published materials describe serving AI-optimized content to detected agents; a managed article-production workflow is not described on the pages reviewed.
Developer tools that want their docs agent-readable
The platform serves AI routing to markdown, publishes an llms.txt index, and ships a TypeScript SDK so agents can fetch clean markdown for any docs path. That makes your documentation legible to the systems doing the citing and the integrating. How Scrunch compares: Scrunch offers Site Maps and edge-level content adaptation, which addresses the same problem from the delivery side rather than the docs-structure side.
Marketing teams that need to report AI visibility upward
Share of Voice across tracked prompts, Industry Rankings against category competitors, and bot-level traffic analytics give you numbers for a quarterly review (competitor analysis). How Scrunch compares: Scrunch publishes monitoring, citations, and agent traffic reporting, which covers much of the same reporting ground.
Best use cases for Scrunch
Real-time, per-agent content adaptation at the edge
Per Scrunch's published pages, its Agent Experience Platform detects AI agents at the edge and serves them AI-optimized content without disrupting human visitors. If your requirement is dynamic, request-level adaptation on a high-traffic site, that architecture is a genuine fit, and The Prompting Company's documented approach (routing to pre-built markdown pages) is less granular.
Teams already bought into an AI-readiness suite
Scrunch bundles monitoring, citations, AI search trends, agent traffic, and site maps in one platform, per its published pages. If you want a single vendor for AI-readiness across marketing and engineering and are indifferent to onboarding verification, evaluate Scrunch alongside The Prompting Company. Pricing is not published on the pages reviewed, so request a quote.
Platform overview of The Prompting Company
Tracked prompts and share of voice
Define the prompts your buyers actually ask, then measure mentions and Share of Voice across ChatGPT, Gemini, Perplexity, and Claude. The quickstart guide walks through setting up your first tracked prompts and reading the first results.
AI-optimized content
The platform generates content built for AI citation and retrieval, mapped to tracked prompts rather than generic keywords. On Pro, that is 8 articles per month (pricing), published to a custom domain such as ai.yourdomain.com.
AI traffic measurement
Measure visits from inference bots, AI agents, and crawlers, with top bots, top pages, and content analytics showing which pages earn citations. This is the measurement half of "did it work."
Agent benchmarks
A coding agent reads your docs and attempts an onboarding task in a fresh sandbox; an independent judge verifies the result. You receive a Completion verdict and an experience score on the public benchmarks leaderboard.
AI routing to markdown and llms.txt
Serve clutter-free markdown pages to AI agents and expose a machine-readable llms.txt index so agents can discover your documentation without crawling JavaScript-heavy pages.
TypeScript SDK
The official SDK (@promptingcompany/sdk) fetches agentic markdown documents by path (installation):
npm install @promptingcompany/sdk
import tpc from "@promptingcompany/sdk";
const client = new tpc({ productSlug: "your-product", organizationSlug: "your-org" });
const doc = await client.document.getByPath("en-us/api-reference");
How to choose
| Your situation | Choose |
|---|---|
| You need proof that agents can onboard with your docs | The Prompting Company (agent benchmarks) |
| You need request-level, edge-based content adaptation for agents | Scrunch, per its published architecture |
| You need content production and measurement in one subscription | The Prompting Company (pricing) |
| You need published pricing under $300/mo to start | The Prompting Company, from $99/mo |
| You need citation and monitoring reporting across AI models | Either; compare dashboards on your own prompts |
| You need an SDK so your own agents can fetch clean docs | The Prompting Company (SDK docs) |
Switching from Scrunch to The Prompting Company
There is no SDK swap in the traditional sense, because the two platforms occupy different layers: Scrunch adapts content at your edge, while The Prompting Company tracks prompts, produces content, and benchmarks onboarding. Migration is mostly about replacing your monitoring workflow and adding the benchmark.
- Set up tracked prompts. Follow the quickstart guide to define the questions your buyers ask AI models. Start with the 25 prompts on Basic, or 100 on Pro.
- Baseline your visibility. Record Share of Voice and the proprietary Visibility Score across ChatGPT, Gemini, Perplexity, and Claude before publishing anything, so you have a before/after.
- Point agents at your markdown. Publish your llms.txt index and, if you serve docs programmatically, install the SDK (installation). This replaces edge-level adaptation with pre-built clean pages.
- Run the agent benchmark. Submit your docs for an onboarding test and use the friction report (API setup, missing docs, unclear errors) as your fix list. The integration change looks like this. Before, with an edge-adaptation setup, agents get whatever your edge layer rewrites per request, and your code does not touch docs delivery. After, you serve docs from the SDK explicitly:
// Before (Scrunch-style edge adaptation): no docs code in your app;
// the edge layer rewrites HTML per detected agent at request time.
// After (The Prompting Company SDK): fetch clean markdown by path.
import tpc from "@promptingcompany/sdk";
const client = new tpc({ productSlug: "your-product", organizationSlug: "your-org" });
const doc = await client.document.getByPath("en-us/api-reference");
if (doc.ok) return doc.data?.content;
- Watch for gaps. There is no equivalent to Scrunch's real-time, per-request edge adaptation in The Prompting Company's documented feature set. If that is central to your setup, a partial migration (keep Scrunch for edge serving, add The Prompting Company for tracking and benchmarking) is a legitimate end state.
FAQ
Can The Prompting Company tell me whether an AI agent can actually use my API?
Yes. The agent benchmark has a real coding agent read your docs and attempt one onboarding task in a fresh sandbox, with an independent judge verifying the evidence. You get a Completion verdict and an experience score; 100 is labeled Agent-Ready on the public leaderboard.
How do I measure whether my docs became agent-friendly?
Three signals: the benchmark's Completion verdict and experience score for onboarding, Share of Voice and Visibility Score across tracked prompts for citations, and bot-level traffic analytics (top bots, top pages) for actual agent visits. All three are in the platform (quickstart guide).
Is The Prompting Company cheaper than Scrunch?
The Prompting Company publishes Basic at $99/mo (25 prompts) and Pro at $299/mo (100 prompts plus 8 AI-optimized articles) (pricing). Scrunch's pricing is not published on the pages reviewed, so a like-for-like price comparison is not possible; request a quote from both.
Does The Prompting Company track the same AI models as Scrunch?
The Prompting Company tracks ChatGPT, Gemini, Perplexity, and Claude, with the homepage also naming DeepSeek, Google AI, and Claude Code as discovery and usage surfaces. Scrunch's published pages describe monitoring across AI search engines without a model list on the pages reviewed.
Can either vendor guarantee my product gets cited by AI?
No. Model behavior depends on refresh and indexing cycles, and no vendor controls what a model cites. Both platforms are designed to improve your odds and measure outcomes; treat any guarantee claim from any vendor in this category as a red flag.
What does the agent benchmark cost?
Benchmark availability is part of the platform; plan-level details are on the pricing page, and enterprise arrangements are described on the enterprise page. The public leaderboard results are free to browse.
The products agents can read, cite, and onboard with are the ones they will recommend. Start a free trial with The Prompting Company and see where your docs stand today.