Tasks and inputs
Qwen3.8 27B is listed for Image-text-to-text, Text generation, Video-text-to-text. It accepts text, image, video and produces text. Use the publisher’s input format and preprocessing for your chosen deployment.
Qwen3.8 27B is an Apache-licensed dense vision-language checkpoint for text, image and video understanding.
The specification score summarizes recorded features and limits. It is not a benchmark result.
Recorded modality support shows what this model can accept and produce. An undocumented modality is shown as unknown rather than assumed to be unsupported.
Accepts
Produces
Qwen3.8 27B is an Apache-licensed dense vision-language checkpoint for text, image and video understanding.
The downloadable model has 27 billion parameters and a native 262,144-token context. The model card describes extending the context to one million tokens with the appropriate serving configuration. Thinking is enabled by default and can be disabled; reasoning effort and preserved thinking are separate controls. A managed version announced by the publisher is described as coming soon, so hosted features should not be assumed to exist in a local checkpoint.
Use the model for an assistant that reads supplied code, documents or visual evidence and returns text. A useful trial includes a repository issue, a diagram question and a video question whose answer depends on a particular moment. Check answers against the supplied material and include cases where the requested fact is missing.
Choose a serving engine that supports the model’s architecture and multimodal processor. Context length, image resolution, video sampling and concurrent requests all affect memory use. Test the native context first before enabling long-context scaling. Preserve the model’s chat template and match the thinking configuration to your application. Downloadable weights do not establish a hosted token price; budget hardware and operations separately. For document extraction, use a fixed output schema in the application and validate extracted values against the original page.
Compare the model on the work you need to complete.
Qwen3.8 27B is listed for Image-text-to-text, Text generation, Video-text-to-text. It accepts text, image, video and produces text. Use the publisher’s input format and preprocessing for your chosen deployment.
Documented limits: 262.14K context. Measure a completed task using the prompts, documents and tools your application will use. Include reasoning and retries in the result.
The weights are downloadable under Apache 2.0. Check runtime support and size the complete model and working context for your expected concurrency.
These records share a directory category. Check the source card, output type and deployment requirements before treating a model as a replacement.
Anthropic
Claude’s small model for high-volume extraction, classification and routing, with adaptive thinking and a 1M-token context window.
Mistral AI
Mistral Large 4 is an open-weight multimodal model released in public preview on the publisher’s announced release, for general reasoning, coding and agent applications.
OpenAI
GPT-6.1 Sol is OpenAI's updated Sol model for coding and professional tasks. It accepts text and images and returns text.
inclusionAI
Ling 3.1 Flash is inclusionAI’s hybrid reasoning model for coding and tool agents, available through a hosted AI Gateway route.
Anthropic
Claude Sonnet 5.5 is an Anthropic language model that accepts text and images and returns text.
MiniMax
MiniMax M3.1 Flash Preview is a multimodal reasoning model for code and agent workflows, accessed through eligible MiniMax plans and tools.