Skip to content

Blog · Published · Updated

A downloadable model: what to review before deployment

Short answer

Downloading a model and permission for a specific way of using it are two separate questions. Before you order an app, describe who uses it, where the model runs and what the customer receives, then check the terms of the specific variant. One family can carry an Apache 2.0 licence, a Llama 3.1 licence and a non-commercial variant at once. An app on a local model costs from €3,000 net at Syntalith.

How we workDetails ↓

  1. 01Call30 minutes with an engineer, free
  2. 02PrototypeA working system on your process and data, free
  3. 03ImplementationTarget and fixed price in the contract; 50% at signing
  4. 04Acceptance50% once the agreed target is met
model on your server or in the cloud in the EU
quoted after the review
same work in application with EU cloud model
from €3,000
task, data and equipment analysis
€0

We belong to the OpenAI and Anthropic partner programmes and learn through their official courses. What this means ↓

Author: Syntalith · 3 min read

Example: the morning request inbox
Agent · ready

Statistics

Open models in companies: statistics 2026

3

different licences in one family of Polish PLLuM models: Apache 2.0, Llama 3.1 and non-commercial. A similar name does not mean the same terms.

Figures checked: 22 October 2026

From an internal tool to a service for customers

A company tested a local model as help for its own employees. Now it wants to offer the feature to customers in a portal. That changes the way the model is used: other users, another product. The team describes this change to the person who assesses the terms, even if the model file stays the same.

Three situations to describe
  • A tool for your own employees

    the model in the company's environment

    a description of the users and where it runs

  • A feature in a customer portal

    the same app, new users

    a fresh assessment of the terms

  • Handing the model files to another company

    distribution

    separate licence terms

A useful description shows who uses the app, where the model runs and what the customer gets: only the output, a feature in the portal, or the model files. The description need not be a full specification. It must tell apart your own app that uses the model from handing its files to another party. The legal assessment belongs to the right person, and a model does not settle it in a chat.

The model card helps you find the material

Hugging Face describes the model card as the place for information on uses, limitations and metadata, including the licence. The person assessing it also needs the full terms the links lead to.

One family, different terms: PLLuM, series 2512
  • PLLuM-4B and 12B

    Apache 2.0

    Commercial use on the terms of that licence.

  • Llama-PLLuM 8B and 70B

    Llama 3.1

    The terms of the Llama 3.1 licence.

  • Variants marked -nc-

    non-commercial

    Research and private uses.

A similar family name does not mean the same terms, so record which model and which release the contractor's proposal covers. Downloading a file from a platform does not replace this check, and after a change of variant you check the licence again. If someone prepared a modified version, you need material on where it came from. The PLLuM variants are described in detail in the article on PLLuM in a company.

The model is not the only component

Other elements work in the app too: document reading, search, speech recognition, libraries and agent tools. According to GitHub (2026), the OpenClaw repository has 392,716 stars and an MIT licence, which shows how popular open agent components are. Each of them has its own terms. In the proposal, include the elements that will really be used, because choosing the language model alone does not close the conversation about the terms of the whole solution.

A list of components to assess
  • The language model: name, release, card address and date checked
  • The document-reading or speech model, if one is used
  • The search engine and the vector database
  • Libraries and agent tools
  • External services, if data can reach them

Alternatives before you sign the order

If the company cannot accept the terms of the chosen model, compare another variant on the same task. The change does not mean giving up the app, but it needs a fresh quality test. An approved external service can be an alternative, if the data requirements allow it. For a simple task, such as copying known fields or finding a document by its number, a search engine or rules are enough, with no text generation. The technical scope should match the user's need, whatever the name of the model from the first trial.

What to agree with the contractor

Syntalith's work includes a list of the components under consideration and a comparison of the permitted variants on the company's task. The client names a person who will assess the terms, and the process owner checks the results. Also agree what happens when a component changes, so the person maintaining the app knows when a fresh assessment is needed. In our private LLM offer we combine choosing the model with a trial on the company's task. Describe who will use the planned app.

This is how we implement it in companies

All case studies →

Client results.

All projects →
foap.comcustomer service

foap.com customer service: 95% of replies are sent automatically.

Users from several countries write to foap.com, in several languages. The system reads the entire conversation and the terms of that person, and then sends a response in that person's language. Questions about money and exceptions go to the team with a ready draft, so they have to check and send them.

the system sends the answers itself, without human typing
95%
time from the arrival of the e-mail to the ready response (median)
30 s
e-mails that require a response receive a ready-made draft
100%
less work for every 1000 answers (estimate)
47 h

Data from foap.com production, October 20-21, 2026. Estimate: 3 minutes for handwritten response.

How foap.com customer service works →
Sample messages. Money matters always come to the person.

Other implementations

jaktako.design

studio projektowe

−87%

Time to look for new customers, every day

4 hours → 30 min

approx. 73 hours less per month (21 days of work)

according to the studio

How the agent loop works →
Signature Estates

real estate

approx. −90%

Operator time spent browsing portals, per day

2–3 hours → 20 min

approx. 45 hours less per month (21 days of work)

respect · contact the seller 5 minutes after the announcement

How the agent finds property listings →
Kuba Koziej, CEO and co-founder of MoreGrowth and board member at Natu.Care

Kuba Koziej

CEO and co-founder, MoreGrowth

„We started with my own AI agent. During the training that followed, I learned to build new applications myself. Today I am bringing this way of working with AI into my business. Working with Syntalith gave me exactly the skills and tools I needed.”

View the original training review on LinkedIn ↗

A model on your server or in the cloud in the EU, proven on your tasks.Book a free process review (30 min) →

How we build automation and agents

From a conversation to a working system in four steps.

  1. 01 · 30 min

    Free process review

    €0

    You're talking to an engineer. You receive the written plan within 2 business days.

  2. 02 · 5 days

    Free prototype

    €0

    The system works on 20-100 cases that your team has already handled. You see every result.

  3. 03 · 2–16 weeks

    Build

    50%

    Fixed price and goal written in the contract. First installment upon signing.

  4. 04 · after reaching the goal

    Reception

    50%

    You pay the second installment when the system reaches its target on your cases.

How a free prototype is created

Mon.

We set a goal

What result do you consider good?

Tue–Wed

The system is working

It goes through your affairs, one by one.

Thu

We compare

We check every difference with the team's work.

Fri

Result and price

List of errors and fixed implementation price.

You choose the matters. We anonymize them upon request, they stay in the EU and we delete them after the test.

On the market, such a test usually costs from PLN 3,000 to PLN 8,000.

Book a free process review (30 min)

We establish a prototype during a free analysis.

Data checked on October 25, 2026.

„Is my data safe?”

Yes. You choose where they go: from your company account to the model on your server.

  1. 00Employee's private account

    Customer and company data ends up in an account that the company cannot see or control. There is no data processing agreement, so it is a violation of the GDPR. When an employee leaves, the chat history goes with them.

  2. 01ChatGPT or Claude corporate account

    The provider does not use your data to train models. A copy of the queries remains in his logs for up to 30 days.

  3. 02API with zero retention (ZDR)

    The provider processes the request and does not save anything. It requires his consent, we prepare the application together.

  4. 03Processing in the EU

    Queries are counted and stored in Europe: OpenAI in the EU region, Claude via clouds with the EU region.

  5. 04Model at your place

    The open model runs on your server or in your cloud in the EU. The data does not go outside the company.

What we do with every implementation

  • Confidentiality agreement and data processing agreement before we receive any data.
  • You choose the cases for the prototype and we anonymize them upon request.
  • The data stays in the EU and we delete it after the test.
  • Corporate accounts with an administrator on your side.

When data cannot leave the company

The open model runs on your server or in your cloud in the EU. To choose from:

White-tailed eagle
Polish model, SpeakLeash and ACK Cyfronet AGH
PLLUM
Polish model, HIVE consortium commissioned by the Ministry of Digitization
Qwen
Alibaba
GLM
Zhipu AI
DeepSeek
DeepSeek

We select a model for the task and check it on your cases. We run it, tune it and maintain it.

Source: Bielik and PLLuM, ACK Cyfronet AGH ↗ · PLLuM, Ministry of Digitization ↗

Do you have a lot of your own data? We will fine-tune the small model.

  1. 1 Your examples
  2. 2 tuning
  3. 3 test on new cases
  4. 4 model at your place

When you have thousands of similar emails, documents or reports, we teach a small open model based on your examples. It works for you, responds quickly and costs less to maintain than a large model. We measure quality on issues that the model has not seen while learning.

Przemysław and Piotr did it in large programming and consulting companies. Artem did it before and does it today at Syntalith. It's about models on your own infrastructure, tuning and quality measurement.

You choose the protection level during the interview. We sign the data processing agreement before we see any data.Schedule a free analysis →

Private AI models

A model on your server or in the cloud in the EU, proven on your tasks.

We select the model and equipment for the task, run it and measure the quality using your examples before you buy anything. The data stays in the company or in the cloud in the EU. We price implementation, hardware or hosting and support separately after analysis.

View private AI models →
model on your server or in the cloud in the EU
quoted after the review
same work in application with EU cloud model
from €3,000
task, data and equipment analysis
€0

Questions

  • Can a downloadable model be used commercially?

    It depends on the licence of the specific variant. In the PLLuM family the 4B and 12B variants carry Apache 2.0, Llama-PLLuM 8B and 70B the Llama 3.1 licence, and the variants marked -nc- are non-commercial. The legal assessment is made by a named person in the company.

  • What changes when customers get access to the model?

    There are other users and another product, so the way the model is used changes. Describe it to the person assessing the terms, even if the model file stays the same.

  • Where do we find a model's licence?

    On the model card. Hugging Face describes the card as the place for information on uses, limitations and metadata, including the licence. The person assessing it also needs the full terms the links lead to.

  • How much does an app on a local model cost?

    At Syntalith from €3,000 net. The proposed work includes a list of components and a comparison of the permitted variants on the company's task. Assessing the terms of use stays with the named person.

Let's talk

Just write, call or choose a date.

We work fast because we hate corporate cotton wool. We don't have salespeople or managers to keep an eye on managers. You talk immediately to engineers who then build your system. We respect each other's time and yours too.

We will accept 3 more companies by the end of the year.

Mon–Fri 9:00–17:00 CET

What our OpenAI and Anthropic partnerships mean for you.

OpenAI makes ChatGPT and Anthropic makes Claude. We belong to the partnership programs of both companies, so we work with their tools every day and learn from the creators themselves.

OpenAI Select PartnerClaude Partner Network
01We will select a tool for the task
We implement both ChatGPT and Claude. We check both on your documents, and you choose the one that makes fewer corrections and saves more time.
02Knowledge straight from the creators
We learn from official OpenAI and Anthropic courses and materials. Your team learns about the features and settings that work today.
03Licenses without our margin
When we roll out ChatGPT, Claude or Copilot, you buy the licences directly from OpenAI, Anthropic or Microsoft and they invoice you. We add nothing to the licence.
04Certified engineers
On the team, we have an AWS certification in machine learning and an Anthropic certification in systems architecture on Claude.

Participation in programs does not mean that OpenAI or Anthropic endorses our services.