Skip to main content
POST
/
v1
/
classify
Classify
curl --request POST \
  --url https://api.classer.ai/v1/classify \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "text": "<string>",
  "image": [
    {}
  ],
  "file": "<string>",
  "labels": [
    "<string>"
  ],
  "classifier": "<string>",
  "descriptions": {},
  "examples": {},
  "text_type": "<string>",
  "priority": "<string>",
  "cache": true
}
'
{
  "label": "technical",
  "confidence": 0.9423,
  "tokens": 142,
  "latency_ms": 87,
  "cached": false
}

​
Request

​
text
string
The text to classify (max 1M tokens). Required unless an image or file is provided.
​
image
string or string[]
Image(s) to classify — URL (https://...) or base64-encoded string. Images above 4MP (2048×2048) are resized. Max 20 images.
​
file
string
PDF or DOCX file — URL or base64-encoded string. Text is extracted and each page is converted to an image for visual analysis. Max 20 pages.
​
labels
string[]
List of possible labels (1–200, max 100 characters each). Required unless classifier is provided.
​
classifier
string
Saved classifier name (e.g. "support-tickets") or a pinned version ("support-tickets@v2"). Alternative to providing labels inline.
​
descriptions
object
Optional descriptions for each label to improve accuracy. Keys are label names, values are description strings.
​
examples
object
Optional per-label example texts to improve accuracy. Keys are label names, values are arrays of example strings.
​
text_type
string
Optional context for prompt tuning (e.g. "app review", "support ticket")
​
priority
string
default:"standard"
Priority level: standard (<1s, 0.20/1M tokens) or `fast` (&lt;200ms, 0.60/1M tokens)
​
cache
boolean
default:"true"
Set to false to bypass cache and force a fresh classification

​
Response

​
label
string
The predicted label
​
confidence
number
Confidence score between 0 and 1 (rounded to 4 decimal places)
​
tokens
number
Total tokens used
​
visual_tokens
number
Number of visual (image) tokens used. Only present when an image is provided.
​
latency_ms
number
Processing time in milliseconds
​
cached
boolean
Whether the response was served from cache

​
Examples

​
Basic classification

import classer

result = classer.classify(
    text="I can't log in to my account",
    labels=["billing", "technical", "sales"]
)

print(result.label)      # "technical"
print(result.confidence) # 0.94

​
With descriptions

Adding descriptions improves accuracy for ambiguous labels: 
result = classer.classify(
    text="We need pricing for 500 users",
    labels=["hot", "warm", "cold"],
    descriptions={
        "hot": "Ready to buy, asking for pricing or demo",
        "warm": "Interested but still exploring",
        "cold": "Just browsing, no clear intent"
    }
)

print(result.label)  # "hot"

{
  "label": "technical",
  "confidence": 0.9423,
  "tokens": 142,
  "latency_ms": 87,
  "cached": false
}