Qwen3.6 35B A3B API

Acerca de 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.

También conocido como Alibaba Cloud Qwen3.6 35B A3B, Qwen3.6-35B-A3B, qwen3-6-35b-a3b

Especificaciones de Qwen3.6 35B A3B

ID del modelo

qwen3-6-35b-a3b

Proveedor

Alibaba Cloud

Categoría

Generacion de texto

Lanzado

29 jul 2026

Ventana de contexto

128K tokens

Salida máxima

16.384 tokens

Entrada

Texto

Salida

Texto

Salida estructurada

JSON Schema

Endpoints

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

IDs de modelo alternativos

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

Precios de la API de Qwen3.6 35B A3BAhorra hasta 72%

Tarifas de pago por uso en vivo del catálogo de EmpirioLabs. Solo pagas por lo que usas, sin mínimo mensual.

Cómo llamar a la API de Qwen3.6 35B A3B

Qwen3.6 35B A3B sirve la API de Chat Completions compatible con OpenAI. Apunta cualquier SDK de OpenAI a https://api.empiriolabs.ai/v1 con tu clave de API de EmpirioLabs y usa el id de modelo qwen3-6-35b-a3b. Consigue una clave de API en el panel de EmpirioLabs.

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)

Parámetros de la API de Qwen3.6 35B A3B

Parámetros de solicitud compatibles con la API de Qwen3.6 35B A3B en EmpirioLabs. Los valores por defecto se aplican cuando se omite un campo.

Información útil

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.

API de Qwen3.6 35B A3B: preguntas frecuentes

¿Cuánto cuesta la API de Qwen3.6 35B A3B?

En EmpirioLabs, Qwen3.6 35B A3B se factura por uso. La tabla de tarifas en vivo de esta página siempre coincide con lo que cobra la API.

¿Cuál es la ventana de contexto de Qwen3.6 35B A3B?

Qwen3.6 35B A3B admite una ventana de contexto de 128K tokens con hasta 16.384 tokens de salida por respuesta.

¿La API de Qwen3.6 35B A3B es compatible con OpenAI?

Sí. Qwen3.6 35B A3B sirve la API de Chat Completions compatible con OpenAI, así que los SDKs de OpenAI existentes funcionan apuntando base_url a https://api.empiriolabs.ai/v1 y usando el id de modelo qwen3-6-35b-a3b.

¿Puedo probar Qwen3.6 35B A3B en el navegador antes de integrar?

Sí. El playground de EmpirioLabs ejecuta Qwen3.6 35B A3B en el navegador con los mismos parámetros que expone la API, para que pruebes prompts antes de escribir código.

¿Cómo consigo una clave de API de Qwen3.6 35B A3B?

Crea una cuenta de EmpirioLabs y genera una clave en API Keys en el panel. La facturación es con créditos de pago por uso, así que solo pagas por las solicitudes que haces.