A new Product Hunt listing called Lettertrace promises to track your AI visibility for free, using your own API keys. The pitch is simple: instead of paying for expensive monitoring dashboards or relying on opaque third-party analytics, you plug in your own OpenAI, Anthropic, or other provider keys, and Lettertrace tells you how often your brand, product, or domain appears in model outputs. It is a small tool with an outsized implication: the AI visibility economy is about to get a lot more transparent, and a lot cheaper.
The timing matters. As of mid-2026, businesses are spending real money trying to figure out whether large language models recommend their products. The phrase “AI visibility” has moved from a niche SEO concern to a board-level metric. Companies want to know if ChatGPT mentions their SaaS tool when a user asks for a project-management solution, or if Claude surfaces their research when a developer asks about a database. The problem has been measurement. Most visibility tracking services charge subscription fees, run their own inference, and return aggregate scores that are hard to verify. Lettertrace’s approach sidesteps all of that by using the user’s own API keys, meaning the data is generated from the exact models and configurations the user cares about, with no middleman markup.
What is genuinely new here is not the concept of tracking AI mentions. Services like Brandwatch and Semrush have dabbled in LLM sentiment analysis. The novelty is the architecture: bring-your-own-keys, zero marginal cost per query beyond what you already pay the model provider, and full control over the prompt templates. That design choice has three consequences worth unpacking.
First, it collapses the cost structure of AI visibility monitoring. A typical enterprise visibility dashboard runs thousands of inference calls per month and charges a premium for the privilege. Lettertrace, by contrast, passes the compute cost directly to the user’s existing API billing. If you already pay OpenAI for your own application’s usage, running a few hundred extra queries to check visibility is nearly free. The tool itself is free, according to the Product Hunt listing. That is a direct challenge to the subscription-based analytics layer that has grown up around the AI economy.
Second, it shifts trust from the vendor to the user. With your own API keys, there is no question about which model version was queried, what temperature was set, or whether the vendor quietly changed its evaluation prompts. Lettertrace’s output is reproducible. You can run the same query set twice, three times, and compare. That reproducibility matters in a market where visibility scores are often treated as black boxes. A founder can now audit the auditor, or skip the auditor entirely.
Third, it exposes a structural tension in the AI economy: the model providers are both the distribution channel and the measurement apparatus. OpenAI, Anthropic, and Google control the APIs that generate the outputs where brands appear or vanish. A tool like Lettertrace, built on top of those APIs, is a thin client over the very systems it measures. That is both its strength and its ceiling. It cannot measure what happens inside closed products like ChatGPT’s free tier, where the user is not billed per token and no API key exists. It only sees what the API sees.
That limitation is worth dwelling on. The free tier of ChatGPT is where most consumer AI visibility happens. A user asks “what is the best CRM for a small team” and the answer may or may not include your product. Lettertrace, using your own API key, cannot observe that conversation. It can only approximate it by running similar queries through the paid API. The approximation is often good, but it is not the same. The same gap applies to Claude’s consumer app and Google’s Gemini. The API is a window, not the whole house.
Still, for the developer-tools and B2B segments, the API is the house. Developers building on top of OpenAI or Anthropic are the ones who care most about whether their product gets mentioned in model outputs, because those mentions drive signups from other developers. For that audience, Lettertrace’s bring-your-own-keys model is not a compromise; it is the correct design. It aligns the measurement with the actual deployment.
The broader implication for AI builders is that visibility is becoming a measurable, auditable asset. Just as search-engine optimization created a whole industry around Google’s ranking algorithm, AI visibility is creating a parallel discipline around LLM outputs. Lettertrace is an early, free tool in that discipline. Its existence signals that the cost of entry for AI visibility monitoring is dropping to near zero, which means the competitive bar shifts. If everyone can measure their AI mentions for free, then having a high mention rate is no longer a differentiator; it becomes table stakes. The differentiation moves to the quality of the product and the strength of the brand, not the sophistication of the tracking.
There is also a policy angle. Tools like Lettertrace make it easier to document systematic bias or omission in model outputs. If a company runs the same query set across multiple providers and finds that one model never mentions its product while another does, that is evidence. In a regulatory environment where the EU AI Act is starting to bite and the FTC is scrutinizing algorithmic fairness, cheap, reproducible visibility data is a powerful instrument. The same tool that helps a startup measure its brand presence can help a regulator measure systemic exclusion.
The catch, and there is always a catch, is that using your own API keys means your queries are visible to the provider. The data you generate about model behavior is also data the model provider sees. There is no anonymity. If you are probing for bias in OpenAI’s outputs, OpenAI knows you are probing. That asymmetry is baked into the architecture and it is worth naming plainly.
Lettertrace is not the first tool to do this, and it will not be the last. But its Product Hunt debut, with the free-forever tag and the bring-your-own-keys framing, lands at a moment when the AI visibility market is consolidating around expensive, opaque enterprise suites. A free, transparent alternative changes the pricing conversation. It forces incumbents to justify their subscription fees against a tool that does the core job for the cost of a few API calls.
What to watch next is whether Lettertrace expands beyond text-based visibility into multimodal outputs, and whether it publishes aggregated, anonymized benchmarks across providers. If it does, it becomes not just a utility but a data source for the entire industry. For now, it is a reminder that in the AI economy, the cheapest way to measure the market is often to ask the market directly, one API call at a time.