Qwen 3.8 27B

Alibaba · 262K token context · multimodal

Pricing: $0.0004 per 1M input tokens, $0.0056 per 1M output tokens

Qwen 3.8 27B is Qwen’s open-weight dense vision-language model for coding, professional workflows, research, multimodal interaction, and long-running agent tasks.

Pricing

$0.0004 per 1M input tokens and $0.0056 per 1M output tokens. Billing is per token: input tokens are charged at the input rate and output tokens at the output rate, from one llmka.dev balance.

Prices checked 2026-11-04

Context and limits

The context window is 262,144 tokens (262K).

The maximum output is 65,536 tokens.

When to use Qwen 3.8 27B

Qwen 3.8 27B (model ID `qwen-3-8-27b`) takes text, image, video input and returns text output.

Supported features in the catalog: function calling (tools), streamed responses over SSE, JSON-schema structured output.

Use Qwen 3.8 27B for text generation and chat.

FAQ

How much does Qwen 3.8 27B cost?

$0.0004 per 1M input tokens and $0.0056 per 1M output tokens. Billing is per token: input tokens are charged at the input rate and output tokens at the output rate, from one llmka.dev balance.

What is the context window and output limit of Qwen 3.8 27B?

The context window is 262,144 tokens (262K). The maximum output is 65,536 tokens.

How do I get access to Qwen 3.8 27B?

Create an API key in the llmka.dev dashboard and send requests to model "qwen-3-8-27b" at https://api.llmka.dev/v1. The endpoint is OpenAI-compatible. One key and one balance cover all models on the gateway.

Links

All models and prices · API documentation · Quick start

Available through the llmka.dev gateway: OpenAI- and Anthropic-compatible endpoints, one key, USDT top-up.