Qwen3.6 35B A3B API

Über Qwen3.6 35B A3B

Qwen3.6 35B A3B is a 256-expert mixture-of-experts reasoning model with 128K context, function tools, and strict structured JSON output.

Text-only. This build does not accept image or video input, unlike the base Qwen3.6 35B A3B. Weights Served from the 2-bit eschamoe W2 build published by Escha Labs (eschalabs.com) as EschaLabs/Qwen3.6-35B-A3B-Escha-W2 on Hugging Face, under Apache-2.0. The experts are quantized to 2 bits, mixed per projection (gate_up_proj at 2-bit and down_proj at 3-bit), the dense layers are int8, and the KV cache is FP16. Escha Labs publishes the quality comparison against an FP8 baseline of the same model: parity or better on math, graduate science, tool use and long context, about 2 percent lower on broad knowledge, and about 7 percent lower on long-horizon code generation, which is the one clear gap. See the model card for the full benchmark table and protocol. Behavior Supports streaming, function tools, structured JSON output including strict schemas, and thinking mode on by default. Set enable_thinking=false for direct answers. With thinking on, the reasoning arrives in reasoning_content and the answer in content, so read both. A low max_tokens with thinking on can be spent entirely on reasoning, so allow room for the answer. Caching Automatic prefix cache reads are billed at the cached-input rate when reported. Explicit cache controls are not supported. Cancelling a streaming request mid-generation bills only the tokens produced up to that point.

Auch bekannt als Alibaba Cloud Qwen3.6 35B A3B, Qwen3.6-35B-A3B, qwen3-6-35b-a3b

Qwen3.6 35B A3B Spezifikationen

Modell-ID

qwen3-6-35b-a3b

Anbieter

Alibaba Cloud

Kategorie

Textgenerierung

Veröffentlicht

29. Juli 2026

Kontextfenster

128K Token

Max. Ausgabe

16.384 Token

Eingabe

Text

Ausgabe

Text

Strukturierte Ausgabe

JSON Schema

Endpoints

POST/v1/chat/completionsPOST/v1/responsesPOST/v1/messagesPOST/v1/completionsPOST/v1beta/models/qwen3-6-35b-a3b:generateContent

Alternative Modell-IDs

qwen3.6-35b-a3bqwen/qwen3.6-35b-a3bEschaLabs/Qwen3.6-35B-A3B-Escha-W2

Qwen3.6 35B A3B API-PreiseSpare bis zu 72%

Live Pay-as-you-go-Preise aus dem EmpirioLabs-Katalog. Du zahlst nur für das, was du nutzt, ohne monatliches Minimum.

So rufst du die Qwen3.6 35B A3B API auf

Qwen3.6 35B A3B bedient die OpenAI-kompatible Chat Completions API. Richte ein beliebiges OpenAI SDK mit deinem EmpirioLabs API-Schlüssel auf https://api.empiriolabs.ai/v1 und verwende die Modell-ID qwen3-6-35b-a3b. Hol dir einen API-Schlüssel im EmpirioLabs Dashboard.

curl https://api.empiriolabs.ai/v1/chat/completions \
  -H "Authorization: Bearer $EMPIRIOLABS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "qwen3-6-35b-a3b",
    "messages": [
      {"role": "user", "content": "Write a haiku about the ocean."}
    ]
  }'

from openai import OpenAI

client = OpenAI(
    base_url="https://api.empiriolabs.ai/v1",
    api_key="YOUR_EMPIRIOLABS_API_KEY",
)

response = client.chat.completions.create(
    model="qwen3-6-35b-a3b",
    messages=[{"role": "user", "content": "Write a haiku about the ocean."}],
)
print(response.choices[0].message.content)

Qwen3.6 35B A3B API-Parameter

Request-Parameter, die die Qwen3.6 35B A3B API auf EmpirioLabs unterstützt. Standardwerte gelten, wenn ein Feld weggelassen wird.

Gut zu wissen

Text-only. This build does not accept image or video input, unlike the base Qwen3.6 35B A3B.

Weights

Served from the 2-bit eschamoe W2 build published by Escha Labs (eschalabs.com) as EschaLabs/Qwen3.6-35B-A3B-Escha-W2 on Hugging Face, under Apache-2.0. The experts are quantized to 2 bits, mixed per projection (gate_up_proj at 2-bit and down_proj at 3-bit), the dense layers are int8, and the KV cache is FP16. Escha Labs publishes the quality comparison against an FP8 baseline of the same model: parity or better on math, graduate science, tool use and long context, about 2 percent lower on broad knowledge, and about 7 percent lower on long-horizon code generation, which is the one clear gap. See the model card for the full benchmark table and protocol.

Behavior

Supports streaming, function tools, structured JSON output including strict schemas, and thinking mode on by default. Set enable_thinking=false for direct answers. With thinking on, the reasoning arrives in reasoning_content and the answer in content, so read both. A low max_tokens with thinking on can be spent entirely on reasoning, so allow room for the answer.

Caching

Automatic prefix cache reads are billed at the cached-input rate when reported. Explicit cache controls are not supported. Cancelling a streaming request mid-generation bills only the tokens produced up to that point.

Qwen3.6 35B A3B API: häufige Fragen

Wie viel kostet die Qwen3.6 35B A3B API?

Auf EmpirioLabs wird Qwen3.6 35B A3B nach Verbrauch abgerechnet. Die Live-Preistabelle auf dieser Seite entspricht immer dem, was die API berechnet.

Wie groß ist das Kontextfenster von Qwen3.6 35B A3B?

Qwen3.6 35B A3B unterstützt ein Kontextfenster von 128K Token mit bis zu 16.384 Ausgabe-Token pro Antwort.

Ist die Qwen3.6 35B A3B API OpenAI-kompatibel?

Ja. Qwen3.6 35B A3B bedient die OpenAI-kompatible Chat Completions API. Bestehende OpenAI SDKs funktionieren, indem du base_url auf https://api.empiriolabs.ai/v1 setzt und als Modell-ID qwen3-6-35b-a3b verwendest.

Kann ich Qwen3.6 35B A3B im Browser testen, bevor ich integriere?

Ja. Der EmpirioLabs Playground führt Qwen3.6 35B A3B im Browser mit denselben Parametern aus, die die API bietet. So kannst du Prompts testen, bevor du Code schreibst.

Wie bekomme ich einen Qwen3.6 35B A3B API-Schlüssel?

Erstelle ein EmpirioLabs-Konto und generiere dann einen Schlüssel unter API Keys im Dashboard. Die Abrechnung erfolgt über Pay-as-you-go-Guthaben, du zahlst also nur für deine Requests.