Tasks and inputs
Hunyuan Hy4 Preview is listed for Text generation. It accepts text and produces text. Use the publisher’s input format and preprocessing for your chosen deployment.
SyncDev model directory · Profile updated Oct 13, 2026
Hunyuan Hy4 Preview is Tencent’s Apache-licensed large MoE text checkpoint with a one-million-token context.
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
Hunyuan Hy4 Preview is Tencent’s Apache-licensed large MoE text checkpoint with a one-million-token context.
The card describes a 770B-parameter backbone with about 49B active parameters and an additional multi-token-prediction component. The auxiliary component is separate from the backbone figures. This is a preview checkpoint for text generation, not a confirmed equivalent of every managed Hunyuan endpoint.
Try a long repository task or an analysis over documents with constraints spread across several sections. Include repeated names and conflicting passages so the trial tests evidence selection rather than simple keyword recall. For an agent, assess the completed artifact and the interaction history, including errors and unnecessary actions. Keep a shorter request in the same evaluation to reveal the ordinary serving experience.
The one-million-token context allowance has substantial implications for KV-cache memory and input processing. Test concurrency and realistic history lengths before allocating production capacity. An active parameter count is not the storage or memory requirement for the complete model. Follow the publisher’s runtime, expert-parallel and prediction-component instructions rather than substituting a recipe for an earlier Hunyuan model.
The downloadable weights use Apache 2.0. An API token price remains a property of a named host, so it is not inferred from the checkpoint. Record the exact revision, template and serving settings in your evaluation to make a later comparison reproducible.
Compare the model on the work you need to complete.
Hunyuan Hy4 Preview is listed for Text generation. It accepts text and produces text. Use the publisher’s input format and preprocessing for your chosen deployment.
Documented limits: 1M 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.