Evaligo
Evaligo is a no-code platform for building AI workflows from visual nodes, running them in batch and deploying them as REST APIs. It also covers prompt engineering: A/B testing of agentic flows, evaluation and production tracing.
What is Evaligo?
Evaligo is a node-based AI workflow platform published by clientcues, presented as a way to connect nodes, process data at scale, and A/B test your agentic AI flows before deploying them as production APIs, with no code required.
Two products sit under one account. AI Workflow Automation is the visual canvas: pre-built nodes cover page scraping, Google search, browser automation, HTTP requests, HTML/CSS and XPath selectors, link extraction, JSON parsing, array operations, text transformation, regular expressions, conditional logic and object manipulation. More than twenty AI models are selectable from a drop-down, with OpenAI, Anthropic's Claude, Google Gemini and Llama named, and Azure OpenAI, Amazon Bedrock and Google Vertex AI documented as provider routes. Users may supply their own OpenAI, Anthropic or Google keys, in which case usage is unlimited and consumes no platform credits. Prompt Engineering is the second half: a playground, experiments, connectable datasets, LLM-as-a-judge and custom evaluators, human annotation, prompt versioning with regression checks wired into CI/CD, production tracing, real-time monitors and dashboards.
The capability the site pushes hardest is built-in A/B testing of agentic flows: two variants run side by side, quality, cost and latency are compared, and the better one is promoted. Deployment is the other emphasis, since a finished flow becomes a REST API in one click, a scheduled job or an embedded widget.
A library of more than a hundred templates is organised by job family, covering SEO, marketing, content, sales, data processing and developer tools across nineteen use-case categories, with dedicated pages for the healthcare, real estate and e-commerce verticals. Flows can also be created by describing the goal and letting the AI assemble them, and a Python SDK is documented for API-key management.
The site claims 2,800 or more users, 4.9 out of 5 from 156 reviews, over a hundred integrations, ten or more AI providers, "80% fewer AI failures" and that "95% see results in their first week"; none of these figures is attached to an identifiable source. What is not covered is equally clear: no mobile application, and no published team or enterprise plan, since collaboration, SSO and enterprise features are announced as still in development.
What it does
- Build an AI workflow by connecting visual nodes, with no code
- Deploy any workflow as a production REST API in one click, with endpoints, authentication, documentation and code samples
- Call more than twenty AI models (GPT-4, Claude, Gemini, Llama) from a drop-down, or plug in your own provider keys
- A/B test two versions of an agentic flow and compare quality, cost and latency before promoting the winner
- Scrape web pages, drive a browser and extract structured data against a JSON schema
- Process hundreds or thousands of items in batch, synchronously or asynchronously, on a one-off or recurring schedule
- Trace and monitor the quality, cost and latency of production runs from dashboards and run history
When to use Evaligo / When not to
A quick filter to help you decide if Evaligo is the right fit.
When to use Evaligo
- Developers and product teams that need to ship AI features to production as REST APIs without managing any infrastructure
- Prompt engineers and AI quality teams running evaluations, LLM-as-a-judge scoring, custom evaluators and human annotation on their prompts
- SEO teams, with six dedicated templates covering keyword research, content briefs, SERP analysis and Google Search Console reporting
- Marketing, content and e-commerce teams, with eleven templates for social posts, ad creative, competitive monitoring and email sequences
- Solo founders, freelancers and non-technical operators, since no code is required and high-volume work can be run in batch or on a schedule
When not to use Evaligo
- Organisations that need a formal compliance file before buying: no DPA, no subprocessor list, no retention period and no declared hosting country are published
- Teams that require shared workspaces, SSO or enterprise administration, all of which are announced as still in development and available only on request from support@evaligo.com
- Anyone who wants a mobile experience: Evaligo runs in the browser only, with no iOS or Android application
- Workloads that need heavy concurrency, since the platform caps concurrent API executions at five on the free plan and ten on Pro
- Buyers whose main criterion is the widest possible catalogue of business connectors: the comparison page itself concedes that Make.com has broader app coverage
How to use Evaligo
A typical end-to-end flow, from setup to results.
- Sign up with an email address alone: no credit card is asked for, and the site promises building can start in seconds
- Browse the template library and click "Try it" on a workflow that matches your need; it is imported automatically into your workspace
- Alternatively, start from an empty canvas in the flow builder, or describe what you want to build and let the AI assemble the flow for you
- Design the flow by dragging and dropping the nodes you need onto the canvas
- Connect the nodes and map the data that passes between them, choosing synchronous, asynchronous or batch execution
- Set up the integrations the flow requires: most templates ask for one to three connections, an AI provider plus an application OAuth such as Google or Slack
- Optionally register your own OpenAI, Anthropic or Google API keys so that model calls run without consuming credits
- Run a test with sample inputs and read the per-node output, the token count and the cost before the real execution
- Adjust prompts and node settings, then deploy in one click as a REST API, schedule the flow on a recurring run, or embed it as a widget
- Follow production behaviour through run history, traces, monitors and dashboards; the site states that most templates deploy in under five minutes
Pros & Cons
Pros
- Very low paid entry point: 7 USD a month for 10,000 credits, cancellable at any time
- Permanent free plan with 500 credits a month and access to all the building features, with no credit card required at sign-up
- The user's own provider API keys are accepted on every plan and consume no credits, keeping AI running costs under control
- A/B testing of agentic flows is included, with no extra tooling to buy or wire in
- Unusual combination under one roof: workflow building on one side, prompt evaluation and production observability on the other
- One-click REST API deployment with no infrastructure to provision or maintain, backed by public structured documentation for both products
- More than a hundred ready-made, editable templates, with cost and token consumption shown before the real run, and a 30-day money-back guarantee advertised on the home page
Cons
- Privacy policy and terms are very generic: no retention period, no subprocessor list, no DPA, no legal basis for processing and no international transfer mechanism
- SOC 2 Type II, ISO 27001 and GDPR compliance are claimed on the security page without a certification body, an audit date or a certificate number
- The publisher is barely identifiable: no postal address, no complete legal entity and no named officer, only the name clientcues in the footer, and the contact page shows no email address, so an account has to be created to write
- Date inconsistencies: the legal documents and the 1.0.0 changelog entry are dated 15 January 2025, while the domain was registered on 17 August 2025 and first archived on 9 October 2025
- Usage figures vary from one block to the next on the same home page, with "2,800+ users" and "2,847+ developers" side by side, and the "Trusted by teams at" logos are generic and unverifiable
- No team plan and no SSO are available, both announced as forthcoming, and concurrency is capped at five API executions on Free and ten on Pro
- Nothing is published about whether customer data feeds model training or about any opt-out, and the integrations page is rendered in JavaScript, so the claimed hundred-plus connectors cannot be read from the page itself
Pricing & Plans
Evaligo offers a permanent free plan, described on the site as "Free forever", which includes 500 credits per month and requires no credit card. The lowest paid entry point is the Pro plan at 7.00 USD per month, which raises the allowance to 10,000 credits per month and can be cancelled at any time; additional credit packs may be purchased on that plan and, according to the site, never expire. Credits are consumed by AI model calls routed through the platform's provider keys, in proportion to the model used and the number of tokens processed; calls made with the user's own OpenAI, Anthropic or Google keys consume no credits, on any plan. When credits run out on the free plan, the AI nodes relying on the platform's keys are paused until the next billing cycle. No team or enterprise pricing is published. A 30-day money-back guarantee is advertised on the home page.
- 500 credits per month
- visual flow builder
- more than twenty AI models (GPT-4
- Claude
- Gemini)
- web scraping and browser automation
- all integrations
- API deployment
- everything in Free plus 10
- 000 credits per month
- the option to buy additional credit packs
- recurring schedules (hourly
- daily
- weekly)
- ten concurrent API executions and priority email support
- cancellable at any time
- not published. Team collaboration
- SSO and enterprise features are announced as in development and are reachable only by contacting support@evaligo.com
Data, GDPR & hosting
A consolidated view of how Evaligo handles your data.
GDPR overview
Compliance here is claimed rather than documented. The security page displays "SOC 2 Type II, GDPR, and industry security standards compliance", with an ISO 27001 badge alongside; none of the three is accompanied by a certification body, an audit date or a certificate number, so none can be verified from the site. The privacy policy, dated 15 January 2025, grants access, correction and deletion rights plus a marketing opt-out: "You have the right to access, update, or delete your personal information." Beyond that it is silent on everything the regulation expects: no legal basis for processing, no retention period, no subprocessor list, no international transfer mechanism, no data protection officer and no Article 27 EU representative. There is no dedicated privacy email either; requests go through the general contact form.
Who owns the data?
Neither the terms of service nor the privacy policy contains an ownership or licence clause covering the workflows, prompts, datasets and content users bring to the platform: the question is simply not settled in writing. What the documents do state is that the account holder is responsible for "Maintaining the confidentiality of your account credentials" and for "All activities that occur under your account", and that the publisher does "not sell, trade, or otherwise transfer your personal information to third parties without your consent, except as described in this policy". Users may ask to access, update or delete their personal information. Running a flow on your own OpenAI, Anthropic or Google key keeps those model calls off the platform's provider keys.
Reuse rights
The privacy policy lists the purposes the publisher claims for the data it collects: providing, maintaining and improving the service, processing transactions and sending related information, sending technical notices and support messages, answering comments and questions, and communicating about products, services and events. Collection is declared at account creation, when contacting support, when subscribing to the newsletter and when taking part in surveys or promotions; marketing communications can be opted out of at any time. Nothing in the documents either authorises or forbids the use of customer content to train models, and no licence-back or restriction is placed on what users do with their own outputs, so nothing in writing limits their reuse. AI calls run either through the platform's provider keys, consuming credits, or through the user's own keys.
Data retention & training
Hosting summary
No hosting country, region or jurisdiction is declared anywhere on Evaligo's site. The security page addresses technical controls only, namely AES-256 encryption in transit and at rest, multi-factor authentication, role-based access control, round-the-clock monitoring, third-party audits and an incident response plan, and never says where data physically sits. The privacy policy and the terms are equally silent: no data centre, no residency commitment, no international transfer mechanism and no subprocessor list. The domain resolves to 143.204.55.92, an Amazon CloudFront address (AS16509) observed from Zurich, but that is a content delivery edge node and says nothing about where workflow data, run history or datasets are stored. The domain is registered with IONOS SE in Germany, which likewise indicates nothing about hosting. An organisation with a data residency requirement therefore has no published information to work from and would have to obtain it directly from the publisher.
Things to keep in mind
Risks and trade-offs to weigh before adopting Evaligo.
- The publisher cannot be identified: no postal address, no legal form and no named officer appear anywhere, only clientcues in the footer, which would make any dispute or formal request difficult to address
- SOC 2 Type II, ISO 27001 and GDPR are claimed without a certification body, an audit date or a certificate number, so they should be read as marketing statements until evidence is produced
- No retention period is published, no subprocessor is listed and no DPA is offered, which leaves the fate of your prompts, datasets and run history entirely undefined once you stop using the service
- Nothing is said, in either direction, about whether customer data is used to train models or about any way to opt out, so feeding confidential material into a workflow is an unquantified bet
- No hosting country or region is declared anywhere, which makes any data residency requirement impossible to satisfy on paper
- Several signals on the site do not hold together: legal documents dated 15 January 2025 predate the 17 August 2025 domain registration, home-page usage figures differ between blocks, the client logos are generic, and "24/7 technical support" is advertised while priority email support is a Pro feature and the contact page carries no email address
- Deploying an agentic flow in one click makes it easy to put unreviewed AI output straight into production; A/B testing and LLM-as-a-judge measure relative performance, not correctness, so skipping the human annotation step risks automating mistakes at scale
Setup & Integrations
Technical difficulty
Low for the essentials. Sign-up needs an email address and no credit card, templates import in one click, and the site states that most deploy in under five minutes. Building a flow requires no code, since nodes are dragged, dropped and connected. Each template still needs one to three connections, an AI provider plus an application OAuth such as Google or Slack, and users must grasp credits, node-to-node data mapping, synchronous versus asynchronous modes and JSON output schemas. Adding your own provider keys, scoped API keys or the Python SDK is optional and clearly aimed at technical profiles.
Deployment
Integrations
Behind Evaligo
Social
Resources
All the official URLs gathered for verification and reference.
Alternatives
Tools that compete with or complement Evaligo.
Frequently asked questions
What exactly is a credit on Evaligo?
Can I use my own AI provider API keys?
What happens when my credits run out?
Do I need to write code to use Evaligo?
Can a workflow be exposed as an API?
Are the templates free?
Is there an Evaligo mobile app?
Is there a team or enterprise plan?
Does Evaligo publish a DPA or a list of subprocessors?
Which security certifications does Evaligo hold?
Should you pick Evaligo?
Evaligo packs an unusual amount into a very small price. The only published paid plan costs 7 USD a month, and the free plan is permanent rather than a countdown, which makes the platform easy to assess at no cost. The combination is uncommon: a visual workflow builder, native A/B testing of agentic flows, and prompt evaluation with production tracing, all under one account. Because the user's own OpenAI, Anthropic or Google keys are accepted on every plan without consuming credits, the running cost stays under control even on the free tier.
The counterpart is transparency. The publisher is identified only by the name clientcues in the footer: no postal address, no complete legal entity, no named officer. The privacy policy and the terms are a handful of generic paragraphs that settle none of the questions a professional buyer asks, namely how long data is kept, who the subprocessors are, whether a DPA exists, where data is hosted and whether customer content feeds model training. The SOC 2 Type II, ISO 27001 and GDPR badges on the security page carry no certifier, no audit date and no certificate number, so they remain claims rather than established facts. The dates do not line up either: the legal documents are stamped 15 January 2025, seven months before the domain was registered.
The tool is young, with a domain registered in August 2025 and a first web archive capture in October 2025. It is well worth trying on non-sensitive material, and its price makes that inexpensive. Before routing regulated or confidential data through it, ask the publisher directly for the retention, hosting, subprocessor and training answers the site does not give.
- Choosing a selection results in a full page refresh.
- Opens in a new window.