Yedric AI Review (2026): Pricing, Features & Honest Verdict

TLDR

Yedric AI is an in-app assistant for SaaS and Shopify apps that runs actions through tool calling instead of only answering questions. It is unusually capable for a free product, but the business model is unstated. Best for: Shopify app developers. Price: from $0/mo plus your own model tokens, free plan yes. Rating: 7/10.

What is Yedric AI?

We found Yedric AI on Product Hunt, where it launched on 1 October 2026 from a team of three makers and was tagged as free. It is an assistant you embed inside your own software. Your users type what they want, and Yedric calls the tools you have exposed to get it done: create a discount code, fix a disconnected integration, change a setting.

The product comes from ShopPad Inc., the company named in the terms and the site footer. ShopPad already runs a family of Shopify apps, and that history shows. The docs, the auth options and several built-in tools are shaped around Shopify app developers first, with general SaaS teams as a second audience.

What Are Yedric AI’s Key Features?

Tool calling with per-action approval

Actions are the core of the product. You connect an MCP server, import an OpenAPI spec, point Yedric at your HTTP API docs so it can generate tools, or write the definitions by hand. Each action then gets one of three modes: Automatic, Ask first, or Off. That per-tool approval setting is the feature that makes letting a model touch production data feel survivable, because destructive actions can always require a confirmation from the user.

Bring your own model, or start on theirs

New assistants start on a small default model that runs on Yedric’s own provider keys, and the organization docs name it as GPT-4o Mini. Add your own OpenAI, Anthropic or Google keys and the model dropdown opens up to whatever those providers offer. The marketing site says there is no platform markup on provider usage, and a per-assistant Usage page reports tokens, estimated cost, cache hit rate and the share of output spent on reasoning.

Secure mode and page context

Secure mode binds each session to a verified user through JWT, signed sessions or Shopify session tokens, so credentials do not sit in the browser. Page Prompts let you set different suggested questions and extra instructions per URL pattern, and a JavaScript API lets your app pass the current plan, shop domain or recent user activity into the conversation as context.

Testing, versioning and analytics

Simulations run saved test scenarios against a changed assistant, with pass and fail criteria and an option to run against a real store with whitelisted tools. Every configuration change is versioned and can be restored or synced to a Git repository. The dashboard tags sentiment and frustration, pulls out feature requests from conversations, and suggests edits to your instructions and documentation gaps.

What Is It Like to Use Yedric AI?

Setup follows a predictable path. You sign in, create an assistant, write instructions, load your docs (a website crawl, PDF, Word or text files, or hand-written entries), configure the widget, then copy a two-line embed snippet into your app. The docs estimate 20 to 30 minutes for a first setup, which looks realistic for the Q and A part. Wiring up real actions against your own endpoints will take a developer longer.

The documentation is better than most launch-day products. It covers webhooks down to retry counts and a 10-second timeout, warns that client tool results are held in memory per Node process, and admits where Trends charts and filtered lists can disagree by a conversation or two because one uses UTC. That candour is a good sign.

The public changelog backs this up. It lists shipped work back to May 2026, with a dense run of additions through August and September: outbound webhooks, usage reports, a mascot builder, human handoff with an agent inbox, and a Yedric MCP server so an outside MCP client can read assistants and conversations. This is a product with sustained engineering behind it, not a weekend demo.

One small surprise: the mascot builder draws and animates a custom character for the widget, and the docs say a full set generally costs USD 3 to 5 in Google AI Studio credits, charged to your own key. It is optional, but it is one of the few places where the docs put a number on what you pay outside Yedric.

Where Does Yedric AI Fall Short?

The biggest gap is the business model. The marketing site says 100% free with unlimited assistants and unlimited conversations. The docs FAQ is softer: free to get started, and for more advanced models or large scale you bring your own provider keys, with a contact address for more details. Nowhere does Yedric say how ShopPad plans to earn money from it, or whether paid tiers are coming.

The terms add to that uncertainty. They describe payments, monthly invoicing in USD, trial periods, a no-refund policy and 30 days of written notice to cancel, and they say all Yedric plans are subject to a fair use policy with unspecified limits. Some of this is likely shared ShopPad boilerplate, but it means the paperwork is ready for paid plans that the site does not show.

Data handling is thin. The privacy policy covers site visitors, account holders and Shopify installs, yet it says nothing specific about the end-user conversations flowing through your assistant, which model providers receive them, how long transcripts are kept, or whether any of it is used to improve the product. If your users discuss orders or account details with the assistant, you will need to ask ShopPad directly.

The social proof is all in-house. All five apps in the logo strip are listed on ShopPad’s own company site, and the quote and case study come from a ShopPad staff member. The case study’s numbers, such as 38% of merchants returning for more than one session over 30 days, are useful, but there is no independent customer on the site yet.

Two smaller points. Provider keys you add can be revealed in the dashboard by an Owner or Admin, unlike tool secrets, which are write-only. And the documentation storage limit is shown in the dashboard rather than published, so you cannot size it before signing up.

How Much Does Yedric AI Cost?

Yedric AI costs $0 per month. There is no pricing page, no paid tier listed anywhere on the site or in the docs, and signup asks for no card. Your real cost is the model usage you pay to OpenAI, Anthropic or Google once you add your own keys, and that scales with conversation volume, tool calls and the model you pick. Yedric publishes no per-conversation estimate, though its Usage page shows estimated spend once you are running.

The default GPT-4o Mini model costs you nothing, since it runs on Yedric’s keys, but the docs frame it as a starting point and want you to bring keys for heavier use. The fair use policy in the terms applies to every plan, and it gives no numbers.

Against the competition, that is a very low entry point. Intercom charges $0.99 each time Fin resolves or routes a conversation, on top of seats that start at $29 per seat per month on annual billing. Chatbase has a free tier with 50 message credits a month, then $40, $150 or $500 per month. CopilotKit has a free Developer plan and Pro at $39 per month. On Yedric the only line item is your provider spend, which for a modest app can come in far below any of those.

Who is Yedric AI Best For?

Use Yedric AI if you build a Shopify app or a SaaS product with a clean HTTP API, and you want an in-app assistant that changes settings and creates records for users rather than linking them to help articles. It fits teams that already pay a model provider and want fine control over which actions run automatically and which ask first.

Use it too if you want to test the idea cheaply. With no fee and a free default model, the cost of a test is mostly developer time.

Skip Yedric AI if you need a signed DPA, a clear retention policy for end-user conversations, or a vendor with a published price you can budget against for three years. Skip it too if you want a full help desk with ticketing and an inbox at its centre; Yedric’s human handoff exists, but it is an add-on to the assistant, not the other way round.

Best Yedric AI Alternatives

Intercom Fin

Fin is the established choice for AI support replies inside a full help desk. It costs $0.99 each time it resolves or routes a conversation, and Intercom’s Essential plan starts at $29 per seat per month billed annually. It is far more mature on ticketing and reporting, but it is built for answering and routing, and it gets expensive at volume.

Chatbase

Chatbase trains a chatbot on your content and adds AI actions on paid tiers. The free plan gives 50 message credits a month, then Hobby is $40 per month, Standard $150 and Pro $500. It is simpler to stand up than Yedric but less focused on acting inside your own product.

CopilotKit

CopilotKit is a developer framework for building in-app copilots with your own UI. The Developer plan is free for one developer, Pro is $39 per month and Team is $100 per seat per month. Pick it if you want to own the whole experience in code rather than configure a hosted assistant.

Final Verdict: Is Yedric AI Worth It?

For Shopify app developers, yes, it is worth a trial run right now. The feature set is deeper than the price suggests: per-action approval, simulations against real stores, versioned configuration, usage reporting and a careful set of docs. Few free products offer that much control over what an assistant is allowed to do.

The rating stops at 7 because the free promise has no stated reason behind it, the privacy policy does not cover end-user conversation data, and every reference customer is a ShopPad app. Build on it, but keep your actions behind your own HTTP API so you can move them to another assistant if the terms change.