~24M records
2013–2026 time range
~907 MB Parquet
Monthly updates
download Download code GitHub lab_profile Data Quality Report
Zenodo · ORCID

info About the dataset

UA Vehicle Registry — Enhanced Quality is a derivative dataset built from the official open vehicle registry of the Ministry of Internal Affairs of Ukraine. The original data is published on data.gov.ua.

The original dataset has numerous quality issues: mixed encodings (Windows-1251 / UTF-8), unstable schema (different column names across months), massive duplicates (~85%), and lack of brand/model normalization.

We fix these issues with an automated data quality pipeline and publish clean, analysis-ready snapshots in Parquet and CSV formats.

verified Quality Improvements

text_format

Encoding

Automatic encoding detection (Windows-1251, UTF-8, KOI8-U) and conversion to unified UTF-8.

table_chart

Schema Stabilization

Unification of column names from different months into a single stable schema.

content_copy

Deduplication

Removal of ~85% duplicates. The original dataset contains repeated records for each month.

123

Type Casting

Dates, numeric fields, categorical variables — cast to correct data types.

directions_car

Brand/Model Normalization

Fixing typos, unifying spelling, mapping to standardized brand and model names.

storage

Parquet Format

Columnar format for efficient analytical queries. ~10x compression compared to CSV.

source Source

Information on vehicles and their owners

Ministry of Internal Affairs of Ukraine

table_chart Data Schema

Complete list of columns in the dataset (v2.0). Extended fields are available via API.

#ColumnTypeExampleDescription
1record_idVARCHARd6179fa48a159e3c…Unique record identifier (SHA-256 hash)
2personVARCHARPOwner classification: P (natural person) or J (legal entity)
3reg_addr_koatuuVARCHAR8036100000KOATUU code of the owner's registered address
4oper_codeINT40Registration operation numeric identifier
5oper_nameVARCHARВТОРИННА РЕЄСТРАЦІЯ ТЗ…Full registration operation name
6oper_car_conditionVARCHARusedVehicle condition at operation
7short_oper_nameVARCHARВторинна реєстраціяShort operation name
8oper_descriptionVARCHARКупив вживане авто…Human-readable operation description
9oper_notesVARCHARЛюдина купила вживане…Additional operation notes
10owner_changeINT1Whether ownership changed (1/0)
11is_importINT0Whether the operation is a vehicle import (1/0)
12is_aidINT0Whether related to humanitarian aid (1/0)
13is_credit_leasingINT0Whether credit/leasing is involved (1/0)
14modification_typeVARCHAR—Modification type
15document_basisVARCHARpurchase_contractDocument basis (purchase_contract, etc.)
16entity_typeVARCHARanyOperation entity type
17is_wartimeINT0Whether performed during wartime (1/0)
18is_digitalINT0Whether it is a digital operation (1/0)
19is_stolen_relatedINT0Whether related to theft (1/0)
20is_insurerINT0Whether the entity is an insurer (1/0)
21is_socialINT0Whether it is a social operation (1/0)
22d_regDATE2021-06-15Date of the registration operation (YYYY-MM-DD)
23dep_codeVARCHAR1330313Service center identifier
24depVARCHARТСЦ 1330Service center name
25dep_regionVARCHARХарківськаService center region
26dep_cityVARCHARХарківService center city
27dep_addressVARCHARвул. КлочківськаService center address
28dep_address_numberVARCHAR228Service center building number
29dep_descriptionVARCHARРеєстрація ТЗ, техоглядService center description
30dep_hours_of_workVARCHARПн-Пт 09:00-18:00Service center working hours
31dep_phonesVARCHAR+380577001234Service center phone numbers
32brandVARCHARToyotaVehicle brand (Latin)
33modelVARCHARCamryVehicle model (Latin)
34generationLIST<VARCHAR>["XV70 facelift 2017"]Model generation
35model_suffixVARCHARXLEModel suffix
36make_yearINT2021Manufacturing year
37model_yearINT2021Model year (may differ from manufacturing year)
38vin_year_codeVARCHARMYear code from VIN (10th position)
39colorVARCHARСірийVehicle body color
40kindVARCHARЛегковийVehicle category (passenger, cargo, etc.)
41bodyVARCHARСеданBody style (sedan, hatchback, etc.)
42possible_body_typesLIST<VARCHAR>["Sedan"]Possible body types for the model
43body_subtypeVARCHAR4-doorBody subtype
44body_modifiersVARCHAR—Body modifiers
45vehicle_classVARCHAR—Vehicle class (A, B, C, D, ...)
46trailer_prefixVARCHAR—Trailer prefix
47purposeVARCHARЗагальнийVehicle intended use
48fuelVARCHARБензинPrimary fuel type
49secondary_fuelVARCHARЕлектроSecondary fuel
50is_gboINT0Whether LPG/CNG is installed (1/0)
51is_hybridINT1Whether it is a hybrid (1/0)
52fuel_notesVARCHARБензин + ЕлектроFuel notes
53possible_fuelsLIST<VARCHAR>["Petrol", "Hybrid"]Possible fuel types for the model
54capacityINT2487Engine displacement (cm³)
55power_kwtINT110Engine power (kW). Present in the source from 2026-05; carried onto earlier registrations by VIN.
56possible_displacementsLIST<INT>[2487, 1998]Possible engine displacements for the model (cm³)
57own_weightINT1570Curb weight (kg)
58total_weightINT2070Gross weight (kg)
59payloadINT500Payload capacity (kg)
60n_reg_new_cyrillicVARCHARАХ4521ВРRegistration plate number (Cyrillic)
61n_reg_latinVARCHARAX4521BPRegistration plate transliterated to Latin
62has_old_cyrillicINT0Whether it has a Soviet-era license plate (1/0)
63plate_typeVARCHARUA_MODERNPlate type (UA_MODERN, etc.)
64region_codeVARCHARAXRegion code from plate number
65is_valid_plateBOOLEANtruePlate format validation result
66validation_errorVARCHAR—Plate validation error description
67region_nameVARCHARХарківськаRegion name by plate code
68needs_moderationBOOLEANfalseWhether moderation is needed
69raw_vinVARCHARJTDKN3DU5M1234567Original VIN from source
70vinVARCHARJTDKN3DU5M1234567Normalized VIN code
71is_valid_vinBOOLEANtrueVIN validation result
72vin_validation_errorVARCHAR—VIN validation error description
73vin_lengthINT17VIN code length
74vin_is_17BOOLEANtrueWhether VIN has standard 17-character length
75vin_has_ioqBOOLEANfalseWhether VIN contains forbidden characters I, O, Q
76vin_wmiVARCHARJTDWMI code (first 3 VIN characters)
77vin_wmi_manufacturerVARCHARToyota Motor CorporationManufacturer by WMI code
78vin_wmi_regionVARCHARAsiaManufacturer region by WMI code
79vin_wmi_knownBOOLEANtrueWhether WMI code is known in the manufacturer database
80possible_transmissionsLIST<VARCHAR>["automatic", "manual"]Possible transmission types for the model
81possible_drivesLIST<VARCHAR>["Front wheel drive"]Possible drive types for the model
82backfill_confidenceVARCHARHIGHConfidence level of recovered VIN for 2013–2020 records
83backfill_stageVARCHARstg1Pipeline stage at which VIN was recovered
84source_schema_versionVARCHARv2_17colSchema version of the source MVS file: empty — legacy (before 2026-05), v2_17col — current.
info

Schema v2.0 — 84 fields. Includes extended operation info, service centers, VIN analysis, body/fuel characteristics, and backfill metadata.

api API Access

REST API for programmatic access to data with extended fields. First 1,000 requests are free.

Vehicle Registry — Request Packages

Credits valid for 12 months from purchase. Field selection discussed individually. API keys are personal and non-transferable.

Free $0
1 000 requests 20 base fields
50K $199
50 000 requests Fields by agreement $3.98 / 1K — save 19%
200K $699
200 000 requests Fields by agreement + bulk $3.50 / 1K — save 29%

Additional APIs

Separate endpoints for quick search and verification. Can be combined with registry packages or purchased separately.

swap_horiz

Plate ↔ VIN Lookup

Find VIN by license plate or license plate by VIN. Current and historical records.

$0.001 / request
Min. package: 10K requests — $10
history

Vehicle History

Full vehicle registration history by VIN or license plate: ownership changes, regions, dates. Online marketplace listings.

$0.02 / request
Min. package: 5K requests — $99
fingerprint

VIN Decode

VIN decoding: manufacturer, model, year, body type, engine, country and assembly plant. Pay only for successful decodes.

$0.01 / request
Min. package: 10K requests — $99

Annual Subscriptions

Fixed price, unlimited requests, full data access. Price depends on the number of extended fields.

diamond

5-Year Subscription

Best value for long-term projects

Basic
$70 / mo
billed $4,200 / 5 years ~30 fields: base + corrected types Save 30%
Standard
$140 / mo
billed $8,400 / 5 years ~50 fields: + VIN, generations, modifications Save 30%
Pro
$280 / mo
billed $16,800 / 5 years 80+ fields: full decode + reports + support Save 30%
  • checkUnlimited requests
  • checkBulk export (Parquet / CSV)
  • checkMonthly data updates
  • checkPrice lock guarantee
  • checkMCP Server included free
  • checkData processing & analysis recommendations
hub

AUTOMOTO MCP Server

Connect AI assistants (Claude, ChatGPT, Cursor) directly to registry data via Model Context Protocol. Available for Business/Enterprise plans or as a separate add-on.

open_in_new MCP Gateway
shield

API keys and downloaded data are personal and may not be transferred, resold, or shared with third parties. Violation of terms results in access termination without refund.

All prices are exclusive of VAT. For custom solutions — contact us.

table_chart Extended Fields

From 27 free base fields to 80+ extended with VIN decoding, technical specs, and EU data.

lock_open Open Data (free)

~27 fields

All data from the Ukrainian MIA database — normalized, deduplicated, and quality-enhanced. VIN codes included only where present in the original dataset.

  • Registration number
  • Make / model (normalized)
  • Year of manufacture
  • Color
  • Fuel type (raw)
  • Engine capacity
  • Own / total weight
  • Registration date
  • Operation code
  • KOATUU code

verified Basic

~30 fields

+ on top of base:

  • Corrected body type
  • Corrected fuel type
  • Corrected drive type
  • KATOTTG (new classifier)
  • VIN validation
  • VIN error correction
  • KOATUU decoded name
  • Service center address

star Standard

~50 fields

+ on top of Basic:

  • Vehicle modification
  • Generation
  • Basic VIN decoding (NHTSA)
  • Transmission type
  • Horsepower / kW
  • Euro emission class
  • Additional tech specs

diamond Pro

80+ fields

+ on top of Standard:

  • Full VIN decoding
  • European vehicle decoding
  • Extended technical specs
  • Country of manufacture + plant
  • All analytical tools
  • 5 custom analytical reports/year
  • Priority support & consulting
  • MCP Server access
  • Recovered VINs for 2013–2020 records

bug_report Known Issues

12 categories of issues in the original MIA dataset (~46.8 GB, 146 CSV files), identified and fixed in the Enhanced Quality version.

Fixed Encoding (mojibake) — Windows-1251 / UTF-8, automatic normalization
Fixed Unstable schema — different column names across years, VIN added in 2021
Fixed Delimiter inconsistency — semicolons and commas in different files
Fixed Massive duplicates — up to 88% repeated records (SHA-256 deduplication)
Fixed Mixed data types — numbers as strings, different date formats
Fixed Deleted leading zeros in KOATUU codes (Excel artifact)
Fixed Brand/model variations — dozens of variants for a single name
Fixed Placeholder values instead of null — невизначено, spaces, dashes, NULL
Fixed Corrupted values — ~107K rows in 2022, ~1K in 2019
Fixed Mixed numeric field formats — different separators, ranges, units
Partial Orphaned KOATUU codes — 100+ codes with no match in registries
Partial Source data anomalies — negative weights, unrealistic years, volumes >50,000 cm³

description Documentation & Links

person Author

Mykhailo Klimnyk

Mykhailo Klimnyk

Head of Data, AUTOMOTO

format_quote How to Cite

Klimnyk, M. (2026). UA Vehicle Registry — Enhanced Quality. https://doi.org/10.5281/zenodo.19099441

When using this data, you must credit Mykhailo Klimnyk and AUTOMOTO.AI as authors of the enhanced version, and MIA Ukraine as the source of the original data.

apps Other AUTOMOTO Products

Ready to work with the data?

Download the free dataset or contact us for custom solutions.

FAQ

What is UA Vehicle Registry — Enhanced Quality?

It is a derivative dataset built from the official open vehicle registry of the Ministry of Internal Affairs of Ukraine. We fix encoding issues, stabilize schema, remove ~85% duplicates, normalize brands and models, and publish clean data in Parquet and CSV.

How many records does the dataset contain?

The dataset contains ~24 million unique vehicle registration records covering 2013–2026. The original MIA dataset contains significantly more due to massive duplicates, which we remove.

What is the difference between free and paid fields?

The free Open Data set contains 27 base fields (plate, make, model, year, color, etc.). Paid tiers add corrected body/fuel types, VIN validation, VIN decoding, vehicle generation, horsepower, Euro class, and up to 80+ fields total.

How does API access work?

The API works on credits or annual subscriptions. First 1,000 requests are free. You can purchase packages from 10K to 200K requests, or an annual subscription with unlimited requests. Additional APIs include plate/VIN lookup, vehicle history, and VIN decoding.

Does the dataset include VIN codes?

The free dataset includes VIN codes only where present in the original MIA data. Paid tiers add recovered VIN codes for 2013–2020 records (missing from the original), VIN validation, error correction, and full decoding.

Can I connect AI assistants to the data?

Yes, via AUTOMOTO MCP Server you can connect Claude, ChatGPT, Cursor and other AI assistants directly to registry data via Model Context Protocol. MCP Server is available for Business/Enterprise plans or as a separate add-on.