LLM
Ling-3.0-flash-VL
Ling-3.0-flash-VLA 124B multimodal MoE model with 5.5B active parameters, image and video understanding, reasoning, and up to 1M-token context.
Model details
View repositoryinclusionAI’s Ling-3.0-flash-VL is a 124B-parameter multimodal reasoning model, built on a hybrid sparse MoE architecture that activates only 5.5B parameters per token. With support for text, image, and video inputs and a context window of up to 1M tokens, it is designed for efficient inference across coding, visual understanding, deep research, and long-running agentic workflows.
Input
1export BASETEN_MODEL_ID="<your-model-id>"
2
3curl -N -X POST \
4 "https://model-${BASETEN_MODEL_ID}.api.baseten.co/environments/production/sync/v1/chat/completions" \
5 -H "Authorization: Bearer $BASETEN_API_KEY" \
6 -H "Content-Type: application/json" \
7 -d '{
8 "model": "inclusionAI/Ling-3.0-flash-VL",
9 "messages": [{
10 "role": "user",
11 "content": [
12 {
13 "type": "image_url",
14 "image_url": {"url": "https://example.com/image.jpg"}
15 },
16 {
17 "type": "text",
18 "text": "Describe what you see in this image."
19 }
20 ]
21 }],
22 "chat_template_kwargs": {"enable_thinking": true},
23 "stream": true,
24 "temperature": 1.0,
25 "top_k": 20,
26 "top_p": 0.95
27 }'
