Alibaba (Qwen) / qwen3.6-35b-a3b
Qwen3.6 35B-A3B stands out for Qwen 3.6 MoE with 35B total and about 3B active parameters and Balances reasoning capability with inference efficiency. It is best suited to cost-aware coding agents and complex general reasoning.
Context
131K
token
Input
$0.30
/ 1M tokens
Output
$1.80
/ 1M tokens
Quality
#54
Elo 1182
Overview
Qwen3.6 35B-A3B stands out for Qwen 3.6 MoE with 35B total and about 3B active parameters and Balances reasoning capability with inference efficiency. It is best suited to cost-aware coding agents and complex general reasoning.
Tool calling
Can call tools or functions, so product flows can connect model output to backend actions.
Streaming
Supports streamed responses for chat, assistants, and interfaces that need visible incremental output.
Model maker
Alibaba (Qwen)
Called through SeaLink's unified account, balance, and OpenAI-compatible API.
Protocol
OpenAI-compatible Chat Completions
Base URL: https://api.sealink.io/v1
Pricing
Light production run
100K input + 25K output
$0.075
Higher-volume run
1M input + 250K output
$0.75
The final charge follows generated usage and the applicable selector, such as resolution, duration, input type, or provider unit.
API
model
qwen3.6-35b-a3bbase_url
https://api.sealink.io/v1Auth
Bearer $SEALINK_API_KEYcurl https://api.sealink.io/v1/chat/completions \-H "Authorization: Bearer $SEALINK_API_KEY" \-H "Content-Type: application/json" \-d '{"model": "qwen3.6-35b-a3b","messages": [{"role": "user", "content": "Hello."}]}'
from openai import OpenAIclient = OpenAI(base_url="https://api.sealink.io/v1", api_key="<your-sealink-key>")resp = client.chat.completions.create(model="qwen3.6-35b-a3b",messages=[{"role": "user", "content": "Hello."}],)print(resp.choices[0].message.content)
import OpenAI from "openai";const client = new OpenAI({baseURL: "https://api.sealink.io/v1",apiKey: process.env.SEALINK_API_KEY,});const resp = await client.chat.completions.create({model: "qwen3.6-35b-a3b",messages: [{ role: "user", content: "Hello." }],});console.log(resp.choices[0].message.content);