inclusionAI: Ling 3.0 Flash VL

Ling 3.0 Flash VL combines visual understanding with a long text context and tool use. It can support document assistants that need to interpret a chart or screenshot before connecting the findings to a larger written task.

Compare models

ling-3.0-flash-vl

AvailableinclusionAIChat

Input / Output: $0.0206 / $0.0618
Context: 256K
Max output: 32K
Modalities: VisionChat
Capabilities: Tool usePrompt Cache

About Ling 3.0 Flash VL

Ling-3.0-flash-VL is an open-weight multimodal model from inclusionAI, built on Ling-3.0-flash. It reads text, images, and video, and brings vision into reasoning and tool use rather than treating it as a separate captioning step. It is a mixture-of-experts model with 124B total and 5.5B active parameters, released under the MIT license, and suits document and screen-reading assistants.

Where it works well

Only 5.5B of its 124B parameters are active per token, which keeps inference light for a model of this size.
Images and video both go in, and VideoRoPE encodes frame order so time is understood.
Tool calling is supported, so a chart or screenshot can lead directly to an action.
A hybrid attention backbone is designed to handle long inputs efficiently, such as many pages or frames in one request.
Weights are public under the MIT license for self-hosting and fine-tuning.

When to choose another model

Thinking mode is on by default, so latency-sensitive calls need to disable it explicitly.
It produces text only; image or audio output requires other models.
Mid-size active parameters mean frontier proprietary models still lead on the hardest reasoning.

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": "ling-3.0-flash-vl",
    "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 spec

per 1M tokens

Official price: Input $0.0206 / Output $0.0618 / Cache read $0.004118
Verified price: —
Discount: —

Prompt cache pricing

Cache read

Official price: $0.004118
Verified price: —
Discount: —

	Official priceper 1M tokens	Verified priceper 1M tokens	Discount
Default spec	Input $0.0206 / Output $0.0618 / Cache read $0.004118	—	—
Prompt cache pricing
Cache read	$0.004118	—	—

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 13, 2026, 4:51 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 Ling 3.0 Flash VL in Console with a prompt ready to edit or send.

Help me try ling-3.0-flash-vl with a short message at /v1/messages. Show the reply, latency, and cost.

Use cases

Best for

Vision

Chart and dashboard reading
Hand it a screenshot of a metrics dashboard and ask which series moved and why the change might matter, then pass the finding on to a workflow.
Video review
Ask questions about a recorded demo or CCTV clip, such as when a step happened, and get answers that respect the order of events.
Interface-driven agents
Let an agent read a web page screenshot, decide which control to use, and call a tool to act, checking the result visually.

Prompt examples

This screenshot shows our billing dashboard. Which month had the sharpest drop, and which line item explains it?

Watch this 90-second product demo and list the steps shown, in order, with the moment each one starts.

From this photo of a whiteboard, extract the action items and owners as JSON, then call create_ticket for each.

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

Is Ling 3.0 Flash VL open-weight?

Yes. inclusionAI publishes the model on Hugging Face under the MIT license, so you can download it, run it locally, or fine-tune it. It uses a custom architecture, so serving needs the code released with the weights.

What inputs does it accept?

Text, images, and video through unified multimodal messages. Output is text. It supports tool calls, so it can act on what it sees as well as describe it.

How big is Ling 3.0 Flash VL?

It has 124B total parameters of which 5.5B are active per token, using a sparse mixture-of-experts design. The 42-layer backbone alternates KDA and Gated MLA attention layers at a 5 to 1 ratio.

Does it think before answering?

Thinking mode is enabled by default, so it reasons before replying. Turn it off through the request parameters when you need fast answers, such as simple extraction from an image.

How much does Ling 3.0 Flash VL cost?

On TokenLab, Ling 3.0 Flash VL costs Input $0.0206 / Output $0.0618 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 Ling 3.0 Flash VL?

Ling 3.0 Flash VL accepts up to 262,144 tokens of context and returns up to 32,768 tokens in one response.

Which endpoint should Ling 3.0 Flash VL use?

Use https://api.tokenlab.sh/v1/messages for Ling 3.0 Flash VL. The request example below shows the matching code shape.

Which operations does Ling 3.0 Flash VL support?

Ling 3.0 Flash VL supports Text to text. Select an operation above to see its endpoint and request example.