Alibaba Cloud: Tongyi Intent Detect v3

Tongyi Intent Detect v3 focuses on identifying what a user is trying to do from an utterance. It is useful for directing a conversation to the right task, service flow, or specialist assistant before a longer response is generated.

Compare models

tongyi-intent-detect-v3

AvailableAlibaba CloudChat

Input / Output: $0.058 / $0.144
Context: 8K
Max output: 4K
Modalities: Chat

About Tongyi Intent Detect v3

Tongyi Intent Detect v3 is an Alibaba model that works out what a user wants from a single utterance and picks the tool that fits, in milliseconds. It is a routing step placed before a larger model, not a chat assistant. It can return the intent label, the chosen tool with arguments, or both, and handles multi-turn clarification of vague requests.

Where it works well

Classifies a user message into your predefined intent categories before any long answer is generated.
Selects a function from a supplied tool list and fills its arguments in a structured format.
Asks for clarification across turns when a request is ambiguous.
Short single-letter intent codes can reduce output and latency.

When to choose another model

It accepts only a small number of tools per request, so large tool catalogs need pre-filtering.
It is not built for open-ended answers, writing or reasoning, and should hand off to a chat model.
Its short window and output cap suit short exchanges, not long histories.

Getting started

Create API key
Create a key in Console, then use it with every model on the platform.

Send your first request

Copy the example for your language and run it against the endpoint.

Text to textMessages API

POST/v1/messages

API format:

curl https://api.tokenlab.sh/v1/messages \
  -H "Content-Type: application/json" \
  -H "x-api-key: sk-xxx" \
  -H "anthropic-version: 2023-06-01" \
  -d '{
    "model": "tongyi-intent-detect-v3",
    "max_tokens": 1024,
    "messages": [
      {"role": "user", "content": "Hello!"}
    ]
  }'

Pricing

The Verified price applies to Verified, which costs less on most models. The Official price is the model maker's published price and applies to the more reliable Official route. Auto bills the route that completes the request.

Default

per 1M tokens

Official price: Input $0.058 / Output $0.144 / Text Input $0.058 / Text Output $0.144
Verified price: —
Discount: —

	Official priceper 1M tokens	Verified priceper 1M tokens	Discount
Default	Input $0.058 / Output $0.144 / Text Input $0.058 / Text Output $0.144	—	—

Usage & activity

Success rate is the share of requests that completed. Latency is how long a full response takes; P95 means 95% of requests finished within that time.

Usage & availability

Last 24 hours

Requests
Success rate
P95 latency
Total tokens

Model performance

Last updated Nov 1, 2026, 8:13 AM UTC

Metrics appear once privacy and data volume thresholds are met.

Data is based on aggregate user requests, excluding status checks.

Open in Console

Open Tongyi Intent Detect v3 in Console with a prompt ready to edit or send.

Help me try tongyi-intent-detect-v3 with a short message at /v1/messages. Show the reply, latency, and cost.

Use cases

Customer-service routing
Decide whether a message is about billing, returns or technical help, then send it to the right flow.
Voice assistant dispatch
Map a spoken command to one of your device functions with the arguments filled in.
Pre-filter for larger models
Run a cheap intent check first and call an expensive model only for requests that need it.
Clarification prompts
Detect when a request lacks a required detail and ask the user before executing an action.

Prompt examples

User said: 'move my appointment to next Tuesday afternoon'. Tools: reschedule_appointment, cancel_appointment, check_availability. Pick the tool and arguments.

Classify this message into one of: refund, shipping_status, product_question, complaint, other. 'Where is my parcel? It was due Monday.'

User said: 'turn it up'. Tools: set_volume, set_brightness. Is the intent clear, or should you ask a question?

This model has conditional pricing. Monthly token totals alone cannot produce a reliable estimate; use the detailed pricing for the request specification, cache, and applicable time window.

FAQ

What does Tongyi Intent Detect v3 do?

It identifies what the user is trying to do and which tool should handle it, returning an intent, a function call with arguments, or both. It is a fast routing component, not a general chatbot.

How many tools can it choose from?

Alibaba's documentation says up to ten tools per request. If you have more, narrow the list with simple rules or a retrieval step first, then let the model choose among the shortlist.

Can it replace a general chat model?

No. It classifies and routes, and then your application or a chat model produces the actual answer. Treat it as the first stage of a pipeline rather than the answering model.

Does it handle multi-turn conversations?

Yes. Alibaba describes multi-round support in which the model can ask for missing details when a request is ambiguous. Pass the earlier turns in the messages so it can resolve references like 'that one'.

How do I parse its output?

Alibaba's output uses tagged sections for the intent, tool call and reply, which you extract with a pattern match. Test the exact format on your own prompts before you build the parser.

How much does Tongyi Intent Detect v3 cost?

On TokenLab, Tongyi Intent Detect v3 costs Input $0.058 / Output $0.144 per 1M tokens. The pricing table above shows the full breakdown. Rates depend on the billing unit, specification, and usage. Compare matching conditions in the model's detailed pricing; a single rate does not determine the total cost.

What are the context window and output limit of Tongyi Intent Detect v3?

Tongyi Intent Detect v3 accepts up to 8,192 tokens of context and returns up to 4,096 tokens in one response.

Which endpoint should Tongyi Intent Detect v3 use?

Use https://api.tokenlab.sh/v1/messages for Tongyi Intent Detect v3. The request example below shows the matching code shape.

Which operations does Tongyi Intent Detect v3 support?

Tongyi Intent Detect v3 supports Text to text. Select an operation above to see its endpoint and request example.