Qwen3.6 35B-A3B

AlibabaMCP Tool CallReasoning
alibaba/qwen3.6-35b-a3bOpen Weights

Open multimodal Qwen MoE for local agents that need vision, audio, and code

Benchmark Pricing$0.38 / 1M in$2.25 / 1M out
Context: 262,144 tokens•Max Output: 65,536 tokens•Released: 2026-04-17

TECHNICAL SPECIFICATIONS

Context Window262,144262.1K tokens
Max Output65,53665.5K tokens
LicenseOpen SourceSelf-hostable
Input Modalities
TextVisionVideo
Function / Tool CallingSupported (MCP Ready)
Structured OutputsGuaranteed Schema

MODEL CONTEXT PROTOCOL (MCP) READINESS

✓ Qwen3.6 35B-A3B natively supports OpenAI / Anthropic format tool calling and is fully compatible with MCP servers and agent loops.

MCP Context Window Recommendations

With a context capacity of 262,144 tokens, this model can handle moderately large MCP tool definitions and multi-turn conversation logs comfortably.

STANDARDIZED BENCHMARKS & EVALS

SWE-Bench Verifiedresolved
73.4

API CALL INVOCATION EXAMPLE

// Example: Calling Qwen3.6 35B-A3B via OpenAI SDK
import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.ALIBABA_API_KEY,
});

const response = await client.chat.completions.create({
  model: "alibaba/qwen3.6-35b-a3b",
  messages: [{ role: "user", content: "Hello! Can you inspect MCP tools?" }],
});

console.log(response.choices[0].message.content);

PRICING BY PROVIDER

15 providers
Alibabaqwen3.6-35b-a3b
$0.38 / $2.25
Cortecsqwen3.6-35b-a3b
$0.17 / $0.56
NanoGPTqwen/qwen3.6-35b-a3b
$0.11 / $0.80$0.06 cache
Scalewayqwen3.6-35b-a3b
$0.28 / $1.68
GreenPTqwen3.6-35b-a3b
$0.34 / $2.05
Umans AI Coding Planumans-qwen3.6-35b-a3b
$0.00 / $0.00$0.00 cache
Novita AIqwen/qwen3.6-35b-a3b
$0.25 / $1.49
302.AIqwen3.6-35b-a3b
$0.28 / $1.70
LLM Gatewaynovita/qwen3.6-35b-a3b
$0.25 / $1.49
QVACqwen3.6-35b-a3b
$0.00 / $0.00
LLMTRqwen3-6-35b
$5.00 / $10.00
Zenifraalibaba/qwen3.6-35b-a3b
$0.19 / $0.48
Boundlessqwen3.6
$0.04 / $0.35$0.02 cache
Thinking MachinesQwen/Qwen3.6-35B-A3B
$0.45 / $1.33$0.09 cache
AIHubMixqwen3.6-35b-a3b
$0.25 / $1.52

Rates in USD per 1M tokens. Rates synchronized from models.dev public catalog.