External report .erf
A form, parameters, data composition schema, and result output. Included in Demo.
LenSA · AI for 1C queries
From a task description to a ready-to-use 1C report
LenSA studies your configuration, builds and runs the query, then creates a connectable report or data processor with a form.
If you are looking for a way to apply AI in 1C: LenSA answers questions about your database — where the data sits, why a report adds up wrong, how to get the selection you need. Knowing the query language is not required.
Actual metadata instead of guesses. A ready file instead of text that still has to be transferred into Configurator by hand.
Works with 1C:Enterprise 8.3.16 or newer, including 8.5, in managed forms mode.
One purchase — yours permanently, with no database limit. The Console works offline; coins are used only by the AI assistant and paid builds.
The main outcome
More than a query. A ready-to-use 1C result
Both builders are included in the free Demo. In the full edition, the AI agent inspects result data — with your confirmation.
Where you hit the wall: two tables plus a calculation between them, or data from different subsystems — the Universal Report no longer covers it, and an external report with a form and parameters is usually treated as a specialist's job. Lensa builds one from your description of the task; assembling the file costs one coin.
Data processor .epf
Result preview and an action for selected rows. Included in Demo.
Validated on your database
Queries run in both editions. In the full edition the agent inspects result rows once you confirm it.
The Console without AI
A strong tool in its own right
AI speeds up the path, but your purchase includes a complete engineering console. Its local tools work without AI and spend no coins.
Editor and query builder
A professional query editor with autocomplete and highlighting, a query builder, parameters, and temporary tables.
Data you can see
Render images directly in query results, inspect the data, and export the outcome to Excel.
A result that looks like a 1C report
Render the result through DCS with a selected layout and familiar 1C formatting instead of a raw table.
Long tasks run in the background
Run heavy queries as background jobs, track their progress, and keep the interface free for other work.
Three steps
From a question to a working solution
The result stays visible at every step: validate it, refine it in a dialog, and only then package it for users.
1
Describe the task
Use plain language, without choosing tables, fields, or query syntax.
2
Validate on your database
Lensa reads the metadata of the running database as you go, with no configuration export.
3
Build the result
Choose a report or a data processor and get a ready form for deployment.
4
Check that it adds up
The query runs against your database and the whole result is visible.
In a batch of several temporary tables you can open the result of each one separately and see exactly where a number comes together — the package navigator is a full edition feature. LenSA explains the query piece by piece: which tables are joined, where the filters come from, what each temporary table contributes.
What is read in your database, and what leaves it
Lensa walks the metadata of the running database as the conversation goes: it reads catalogs, registers and their structure exactly where the task leads. No configuration export is needed, and its knowledge never goes stale — change the database and it sees it at once. What travels outward is the query text and the names of metadata objects, not the contents of your database.
Rows from your database are sent only on separate confirmation: you see the exact query text before it runs, and declining does not break the conversation. The free Demo edition has no such capability at all.
Personal data and the details — in the answer on what is sent
Interface
An external report from the discovered query
Parameters, groups, resources, and the form are assembled inside the Console. The result is a file ready to connect to your configuration.
Turn the query into an external report — groupings, totals, and images are set up right here.
From task to finished report — seven steps
- Describe the task in plain words — LenSA walks your database metadata itself: it reads catalogs and registers until the query comes together. On 1C:UT 11 the walk fit into 21K tokens of context.
- Paste a finished query — LenSA breaks it down packet by packet, names the main risk, and rates the severity of every finding.
- Set up execution once — send the result through DCS or run it in the background.
- Open a temporary table and immediately see how long it took and how many rows it produced.
- Refine the query by hand while the editor suggests 1C syntax and configuration metadata.
- Turn the query into an external report — groupings, totals, and images are set up right here.
- Connect your key and proxy, then check updates and the change history in the same view.
Worked example
A complete workflow, not a list of promises
See how LenSA explores a configuration, builds and fixes the query, and turns the result into a ready tool.
- Works with metadata from your actual database
- Explains a query part by part — which tables join, where the filters come from, what each temporary table contributes — yours or someone else's
- AI query correction and precise refinements in a dialog
- Delivers the result to the editor, Excel, a data composition schema, or an external report
- The AI model gets the query text and metadata object names; result rows only with your confirmation
Who it is for
One tool for different 1C roles
The Console does not replace professional knowledge where it matters. It removes the manual work between a question, a query, and a usable result.
1C analyst
Builds reports on sales, stock, warehouses and receipts without waiting for a developer.
1C developer
Writes, validates, and optimizes complex queries faster.
Support specialist
Understands an unfamiliar customer database through its metadata.
Beginning specialist
Solves an applied task without knowing the 1C query language.
When the decision is not yours to make
The comparison you can show a manager sits on its own page: what a finished task costs with LenSA, and what the same work costs when contracted out. The figures are ours, measured on a real session, not market estimates.
Task economics in figuresPath to a result
What changes with LenSA
LenSA does not ask an AI model to guess the internal structure of a database: it supplies real metadata and guides the query to a verifiable result.Traditional console
The query is written and debugged manually.
The result is a data table.AI query console
AI helps create and correct the query text.
The result is a working query.LenSA
AI uses exact object and field names from your configuration together with proven join patterns. LenSA runs the query and validates the result on the database.
The result is a cleaner, better-structured query, a ready report, or a data processor.Product editions
Demo or full version
The Demo is free and never expires. Both editions run the query against your database and show the complete result. In the full edition, the AI agent inspects result data — with your confirmation.Demo edition build 2026.09.07.2
Free
Download and use with no expiry date
500 trial AI coins after signup
One-step signup, no card required · The Console also works without coins
what the demo includes Collapse
- A complete query console: builder, parameters, autosave, query files
- The query runs against your database and the complete result remains visible
- AI agent: metadata and applying edits
- AI models in the Demo: GLM 5.2, Muse Spark 1.2, DeepSeek v4 Pro, Gemini 3.7 Flash, MiniMax M3, and Gemma 4. The roster may change
Full edition build 2026.09.07.2
-50% 5,000 ₽≈ $58 10,000 ₽
One payment, forever
5,000 AI coins included
Launch price. It will rise to ₽10,000 as demand grows
in addition to the demo Collapse
- Professional query editor with autocomplete and syntax highlighting
- Package and temporary table navigator
- Background execution
- The AI agent inspects result data — with your confirmation
- Advanced and flagship AI models
Detailed Demo and full-version comparison
| Feature | Demo | Full version |
|---|---|---|
| AI agent: metadata and applying edits | Yes | Yes |
| Run queries against the database and inspect the result | Yes | Yes |
| Instant query syntax check | Yes | Yes |
| The AI agent inspects result data — with your confirmation | No | Yes |
| Available AI models |
| All available AI models |
| Additional AI models, including advanced and flagship options | No | Yes |
| Dialogue context limit | 120,000 tokens | 200,000 tokens |
| External report .erf builder | Yes | Yes |
| External data processor .epf builder | Yes | Yes |
| Standard 1C query builder | Yes | Yes |
| Query editor and conversation history | Yes | Yes |
| Professional query editor with autocomplete and syntax highlighting | No | Yes |
| Package and temporary table navigator | No | Yes |
| Query parameters | Yes | Yes |
| Result row limit | Yes | Yes |
| Autosave | Yes | Yes |
| Background execution | No | Yes |
| Query text normalization | Yes | Yes |
| Text and executable code for Configurator | Yes | Yes |
| Conversation history and selection | Yes | Yes |
| .q1cs and .q1c query files | Yes | Yes |
| Additional interface themes | No | Yes |
Clear economics
What AI work costs
Actual spend depends on the scope of the task, the selected AI model, and the amount of context used.
A purchase brings 5,000 coins
And a whole task: a finished report with a form and a data composition schema A measured session from a real user: one evening from the brief to the finished file, and the figure covers every follow-up turn of the conversation. How many you will need depends on the task and your experience — which is why we name the final bill rather than a number of attempts.
about $17 Step-by-step breakdown
Coins are charged per request to the AI model, including follow-up turns in the conversation — "add the manufacturer", "drop that filter" — and per finished file built: an external data processor or report. A build is not a neural network, it is server-side work priced per file. A network failure, a missing model response and a failed build all charge nothing. The cost of every turn shows in the conversation right away, not as a total at the end.
You pick the model
LenSA is not tied to a single neural network: the catalogue holds models of different tiers, from basic to flagship, and you can switch between turns of the same conversation. The price gap between the lowest and the highest tier is measured in multiples, and that lever is yours. No provider keys, accounts or subscriptions of your own: one balance, paid in coins.
The list holds only networks that passed a run on basic 1C tasks: one that still fails the task on the second attempt does not make it into the catalogue.
If something does not work Write to support@digitalmechanics.dev. Requests are accepted both by email and through the feedback form — all of them are taken into work. A reply goes to those who wrote by email, within 5 business days.