Configuration

Select your API key and environment to authenticate playground requests.

Developer Settings

Manage global preferences for your API responses.

Universal Price Adjustment

Automatically increment or decrement all price estimations rendered by the API. This is applied proportionally to itemized breakdowns.

Predict by VIN

Get a repair cost prediction using a vehicle's VIN, zip code, and problem identifier.

POST /api/v1/predict
First 11 characters (VIN mask) used for vehicle lookup.
Request
POST /api/v1/predict

{}
Response
// Submit the form to see the prediction result

Predict by Vehicle Specs

Get a repair cost prediction by selecting year, make, model, trim, and engine - no VIN required.

POST /api/v1/predict/by-specs
Request
POST /api/v1/predict/by-specs

{}
Response
// Select vehicle specs to get a prediction

Plate to VIN Lookup

Resolve a US license plate to a VIN, then automatically create/enrich the vehicle in the system.

POST /api/v1/predict/plate-to-vin
Request
POST /api/v1/predict/plate-to-vin

{}
Response
// Enter a plate number to look up its VIN

VIN Lookup

Decode a VIN to get vehicle details (year, make, model, etc.) without running a prediction.

POST /api/v1/predict/vehicle-lookup
Decodes via NHTSA and returns vehicle information.
Request
POST /api/v1/predict/vehicle-lookup

{}
Response
// Enter a VIN to decode vehicle details

Extract VIN from Image

Upload a photo of a VIN plate, sticker, or registration document and let AI extract the VIN automatically.

POST /api/v1/predict/extract-vin

Drag & drop an image here or click to browse

Supports JPEG, PNG, WebP (max 10MB)
Request
POST /api/v1/predict/extract-vin

Content-Type: multipart/form-data

{}
Response
// Upload an image to extract VIN

Search Sub-categories

Search across all visible repair sub-categories by name. Returns matches with their parent category and ID.

POST /api/v1/predict/categories/search
Minimum 2 characters. Returns up to 20 results.
Request
POST /api/v1/predict/categories/search

{}
Response
// Type a search query to find sub-categories

Describe Sub-category

Get an AI-generated description of what a repair sub-category involves. Descriptions are cached after first generation.

POST /api/v1/predict/categories/{id}/describe
Request
POST /api/v1/predict/categories/{id}/describe
Response
// Enter a sub-category ID to get its description

AI Repair Recommendation

Describe what's wrong with the vehicle and get AI-powered repair suggestions matched to our database.

POST /api/v1/predict/diagnose
Minimum 10 characters. Be as descriptive as possible.
Request
POST /api/v1/predict/diagnose

{}
Response
// Describe a car problem to get diagnosis

AI Diagnostic Chat

Conversational AI diagnostic — describe symptoms and answer follow-up questions to get an accurate diagnosis.

POST /api/v1/predict/diagnose/start
POST /api/v1/predict/diagnose/reply
POST /api/v1/predict/diagnose/reset
GET /api/v1/predict/diagnose/{session_id}
How it works:
  1. Start — Send a symptom description (+ optional VIN or Image) → AI identifies vehicle & asks a follow-up
  2. Reply — Answer the question (can attach images) → AI asks another or gives diagnosis
  3. Confirmation — When vehicle is identified, the response message will include a formatted spec summary.
  4. Diagnosis — AI returns ranked repairs with confidence scores matched to our database
  5. Retrieve/Reset — GET to fetch full history; POST /reset to clear conversation & vehicle context.
Cost: 1 Gemini API call per turn. Typical session = 4-5 calls. Only the start counts against your quota — replies within a session are free.
1. Start Session
POST /api/v1/predict/diagnose/start
Request Body:
{
"description": "My car makes a grinding noise" // required
"vin": "1FMCU0MN9PU09186" // optional, 17 chars
"image": BinaryData // optional, multipart/form-data
}
Response:
{
"success": true,
"session_id": "uuid-here", // use this for replies
"message": "What kind of noise is it? Grinding, squealing, or clicking?",
"step": 1, "max_steps": 5, "is_complete": false
}
2. Reply
POST /api/v1/predict/diagnose/reply
Request Body:
{
"session_id": "uuid-from-start" // required
"answer": "It is a grinding noise" // required
"image": BinaryData // optional, multipart/form-data
}
Response (follow-up):
{
"message": "Does it happen only at low speeds or also at highway speeds?",
"step": 2, "is_complete": false
}
Response (final diagnosis):
{
"message": "Based on our conversation, here is my diagnosis...",
"is_complete": true,
"diagnosis": [{ "category": "Brake", "subcategory": "Brake Pad Replacement", "confidence": 0.92, "problem_id": 38 }]
}
3. Reset Session
POST /api/v1/predict/diagnose/reset
Request Body: {"session_id": "uuid"}. Clears history and identification.
4. Retrieve Session
GET /api/v1/predict/diagnose/{session_id}
Returns full conversation history + diagnosis.
Try it — Interactive Chat
API Response
// Start a diagnostic session to begin

Submit Prediction Feedback

Use a session_id from a prediction to submit accuracy feedback.

POST /api/v1/predict/feedback
Request
POST /api/v1/predict/feedback

{}
Response
// Submit feedback to see the result

Repair Lifecycle Prediction

Find out when a specific repair will be needed again, based on historical repeat-repair data for your vehicle's make.

POST /api/v1/predict/repair-lifecycle
Request
POST /api/v1/predict/repair-lifecycle

{}
Response
// Submit the form to see lifecycle prediction

Repair Sequence Prediction

Discover what repairs typically come next after a specific repair, based on historical sequence mining data.

POST /api/v1/predict/repair-sequence
Request
POST /api/v1/predict/repair-sequence

{}
Response
// Submit the form to see what breaks next

Prediction History

All prediction records. Red flag rows are highlighted.

Date Session VIN Problem Local Avg AI Avg Match Memory Feedback Reason
Loading...

Lifecycle Prediction History

All lifecycle prediction records. Red flag rows are highlighted.

Date Session Vehicle Repair Source Local Days AI Days Local Miles Next Due Flag
Loading...

Your API Keys

Manage your API keys for authentication.