At a glance
- Tags
- GDPR, ISO 27001, SOC 2
GPT-trainer in our research
All researchStudy
AI Tool Pricing Index, November 2026
Price in Customer service agents
$150a month
Pricier than 10 of the other 11 tools with a monthly price.
Edition price, checked 5 November 2026.
Pricing pages of the AI tools, November 2026 editionas of n = 12 tools with a monthly price
A multi-agent builder for support and lead capture
GPT-trainer builds voice and text agents that answer from documents you upload and then deploys them on phone calls, email, SMS, web chat and social channels. The homepage describes a setup where an "AI supervisor with intent-based routing" picks which specialized agent handles each message, while background agents label conversations and alert staff.
Agents can call functions and outside MCP servers, collect lead details into variables, and hand a call or chat to a person. The supervisor overrides guide shows how rules can force a specific agent when conditions are met, which turns the routing into something closer to a decision tree when you need it. The voice side adds batch outbound calling, rotating number sets and templated follow-up sequences, so the product also covers outbound sales development as well as inbound support.
Automatic retraining when a website changes is marked No on Self-Serve and Yes on Scale in the pricing table.
Two published plans, with most integrations on the second
Enterprise is "Talk to Sales" and adds unlimited credits, on-prem hosting, an SLA, forward-deployed engineers and source code availability.
The gap between the two self-serve plans is wider than the price suggests. Self-Serve includes 12,000 message credits, 2 chatbots, 4 collaborators, one agent per chatbot and one function call per agent, and limits models to 16k context. HubSpot, Zendesk, SMS, Google Search, image uploads, user identity verification and the AI Supervisor itself are all marked No on Self-Serve and Yes on Scale. A team that wants the multi-agent routing the homepage leads with has to buy Scale.
Usage is metered in message credits. The BYOK guide estimates "Each Message Credit costs ~$0.0032 USD" and lists credits consumed per query by model, from fractions of a cent on GPT-4o-mini to $0.72 on a 64k GPT-4 query. Once credits run out, accounts with their own OpenAI, Anthropic or Google key are charged against that key.
The vendor's pages disagree on models, credits and certification
Several pages carry older claims that conflict with newer ones. The homepage says it supports models from "OpenAI, Anthropic, Google, Mistral, and DeepSeek", while the FAQ still says "Currently, we have GPT-3.5-turbo and GPT-4 as options" and that other providers will be added "when they become more mature". The BYOK guide names OpenAI, Anthropic and Google only.
On credits, the BYOK guide says the platform runs "Rather than having Message Credits (MCs) baked into existing subscriptions", yet the pricing table lists 12,000 and 60,000 credits a month inside each plan.
On certification, the homepage calls the product "fully compliant with SOC II, ISO 27001, and GDPR", while the FAQ says the company is "in the process of securing ISO 27001 certification". The pricing page also offers a "2 weeks Pro trial" although no plan on the page is called Pro.
Security documents sit behind a trust center request
The trust center, hosted on SafeBase, lists GDPR, ISO/IEC 27001:2022 and SOC 2 Type 1, with the SOC 2 report, a pentest report and a network diagram available on request. It also states a recovery time objective of 24 to 48 hours. Its Product Security panel lists audit logging, multi-factor authentication and role-based access control as headings with no description behind them, so those features are not documented anywhere public.
The FAQ is candid on access: "Our engineers have access to documents you upload" for debugging, and stored data is not end-to-end encrypted. The privacy policy places servers "primarily located in Oregon, United States of America" and states that customer data is not used to train models. EU dedicated servers, an SLA and a DPA are described on the homepage as enterprise offerings.
Operated by Paladin Max, the company behind Petal
The FAQ ties the team to petal.org, a reference manager and document analytics product it says has run for four years without a data breach.
The company also runs an invite-only Ambassador program on the pricing page that lets partners resell chatbots built on GPT-trainer for "40% revenue sharing", plus a white-label setup documented in the guide.
Where it fits and where it does not
The product suits a support or sales team that wants one vendor for website chat, messaging channels and phone agents, and is willing to configure agents and rules itself. The documented API covers creating chatbots, agents, sessions and data sources, which makes it usable as a backend for a custom app or a white-label resale.
Self-Serve is restrictive for anything beyond a single bot: no CRM or helpdesk connectors, no AI Supervisor and one function call per agent. EU data residency and an uptime commitment both sit on the quote-only Enterprise tier.
GPT-trainer price and terms
Terms
- Free to start
- PartlyLimited free useFree signup with a 2 week Pro trial and no credit card required
- Card to start
- YesNot required
- Sign up without sales
- YesYes
The terms in full
the limits of a free account after the trial are not published.
For AI agents
- API
- Yesguide.gpt-trainer.com,
What needs a higher plan
- Enterprise
- Data stays in your region, Self-hosted, Uptime promise

Pricing page, captured
gpt-trainer.com/pricing, checked
What we checked on GPT-trainer
What the price counts
Flat monthly subscription tiers, each bundling a fixed pool of message credits per month; Enterprise is quote-only and extra AI usage can run on your own API key.
Self-Serve $150 USD Billed monthly General Message credits 12000 MCs / month
gpt-trainer.com/pricingChecked
Free to use
Signup is free and comes with a two week Pro trial; the site does not describe what a standing free account can do once that trial ends.
Start for free 2 weeks Pro trial Pay as you go Choose your own subscription Public API
gpt-trainer.com/pricingChecked
Self-serve signup
Self-Serve and Scale plans are bought directly in the app with a published price; only the Enterprise tier routes buyers to a demo with sales.
Do It Yourself The perfect plan if you're just getting started with our product. Start for free
gpt-trainer.com/pricingChecked
Card to start
The homepage signup call to action states that no credit card is needed to start, which covers the free start and the trial period.
Start for free No credit card required
gpt-trainer.comChecked
Contract
Both paid self-serve plans can be billed month to month, with an annual option offering two months free; Enterprise terms are not published.
Payment frequency Monthly Annually 2 months free
gpt-trainer.com/pricingChecked
Cancel any time
The terms of service let you cancel your account at any time; they do not describe a cancellation fee, though they also do not describe refunds for unused time.
You have the right to cancel your Account at any time.
gpt-trainer.com/termsChecked
Works with
Slack, Zapier, Make, WhatsApp, Instagram and Messenger are on Self-Serve; HubSpot, Zendesk and SMS need the Scale plan, and Enterprise integrations go through sales.
On the page: Feature comparison table, Self-Serve column, Deploy / Integrations rows
gpt-trainer.com/pricingChecked
Public API
A documented REST API covers chatbots, agents, sessions, messages and data sources, and the pricing table marks API access as included on every plan.
Use GPT-Trainer API for External Applications: Integrate AI into your own app or workflow through GPT-trainer’s API.
Acts on its own
Agents answer routine conversations and run actions such as calendar booking without a person approving each step, escalating to staff on configured criteria.
Incorporate MCP servers and function calls to pull real-time data on-demand or perform common tasks like calendar booking autonomously.
gpt-trainer.comChecked
Limits on what it may do
Rule-based supervisor overrides force a chosen agent when conditions hold, and escalation criteria and working hours set when a person is pulled in.
These conditions, if satisfied, will bypass the AI Supervisor’s own judgment and force a designated Agent to handle the current input query.
No code needed
Agents are built in a no-code builder with document uploads and templates; the API is optional for teams that want custom connectors.
Use our no-code builder or API to create intelligent support agents.
gpt-trainer.comChecked
No training on your data
The privacy policy states the company runs no models of its own and does not use customer data for model training, relying on retrieval instead.
we do NOT use any data we collect from you for training LLMs.
gpt-trainer.com/privacyChecked
Retention you control
Data is kept while the account is active and deleted on a verifiable request; the API also exposes delete endpoints for sessions, messages and sources, but no automatic retention window is described.
on receipt of a verifiable request we will delete your personal information from our records.
gpt-trainer.com/privacyChecked
Data stays in your region
Standard hosting is AWS in Oregon, USA; dedicated EU servers are offered, and the pricing page places dedicated hosting on the Enterprise tier.
We offer dedicated servers for clients in the EU, as well as SLA and DPA for enterprise clients.
gpt-trainer.comChecked
Self-hosted
On-premises or dedicated hosting is offered only on the quote-only Enterprise tier; self-serve plans run on the vendor's AWS cloud.
Dedicated or on-prem hosting with scalable infrastructure
gpt-trainer.com/pricingChecked
Take your data out
The privacy policy promises a portable copy of your personal information on request, and the API can fetch session histories as text transcripts.
Upon a verifiable request, we will deliver a copy of your personal information to you or another entity free of charge and in a readily useable…
gpt-trainer.com/privacyChecked
Security audits
The SafeBase trust center lists SOC 2 Type 1, ISO/IEC 27001:2022 and GDPR, with the SOC 2 report available on request; no ISO 42001 or HIPAA agreement is named.
On the page: Trust Center overview, Compliance panel
security.gpt-trainer.comChecked
Uptime promise
A service level agreement is marked No on Self-Serve and Scale and Yes on Enterprise; no uptime percentage is published.
Service Level Agreement (SLA) No
gpt-trainer.com/pricingChecked
GPT-trainer alternatives
Questions about GPT-trainer
- How much does GPT-trainer cost?
- GPT-trainer costs $150 a month. From gpt-trainer.com/pricing, checked 5 November 2026.
- Does GPT-trainer have an API or an MCP server?
- GPT-trainer has an API. From guide.gpt-trainer.com/introduction, checked 5 November 2026.
- Is GPT-trainer open source, and can you host it yourself?
- You can host GPT-trainer yourself on Enterprise. From gpt-trainer.com/pricing, checked 5 November 2026.
- What have we checked on GPT-trainer?
- 18 facts, under these headings: What it costs to start, what it works with, what it does on its own, what happens to your data, whether a company can buy it.
- How do we check GPT-trainer?
- We hold 18 facts on GPT-trainer, quoted from 7 pages. Facts and price last checked 5 November 2026. Last change found 12 November 2026.
GPT-trainer price history
We have read the price once so far, on 5 November 2026: from $150 a month.
| Price | Plan | First seen | Last seen | Readings | Source |
|---|---|---|---|---|---|
| $150 a month | 5 November 2026 | 5 November 2026 | 1 | Pricing page, read 5 November 2026 |
From our research on 5 November 2026. A date is when we read the page, not when the price changed.
Is GPT-trainer yours?
Free, and the page then carries the verified owner mark. You confirm it with an email at gpt-trainer.com. Claim this listing
Report an error if a fact is wrong.