# Cola-DLM [Canonical model page](https://headlessthrones.com/models/bytedance-seed-cola-dlm) · [Model library](https://headlessthrones.com/models) Public catalog reference from Headless Thrones. Exact catalog ID: `bytedance-seed-cola-dlm`. Catalog source-check date: 2026-10-23. This brief contains the public catalog answers and any catalog guide, not private drafts, historical editorial revisions or independently measured rankings. The canonical HTML page may also display separately published editorial. Current catalog answers ## What to know about Cola-DLM AI-compiled from linked model documentation. This is a catalog reference, not an independent performance review. Catalog checked October 23, 2026. [Exact model](https://headlessthrones.com/models/bytedance-seed-cola-dlm#model-answer-identity) · [Access](https://headlessthrones.com/models/bytedance-seed-cola-dlm#model-answer-access) · [Inputs & outputs](https://headlessthrones.com/models/bytedance-seed-cola-dlm#model-answer-modalities) · [License](https://headlessthrones.com/models/bytedance-seed-cola-dlm#model-answer-license) · [Model details](https://headlessthrones.com/models/bytedance-seed-cola-dlm#model-answer-research) · [Weight details](https://headlessthrones.com/models/bytedance-seed-cola-dlm#model-answer-repository) ### Which Cola-DLM does this page describe? Cola-DLM is a model record from **ByteDance Seed** in the Cola family. [Source](https://huggingface.co/ByteDance-Seed/Cola-DLM) The recorded model identifier or checkpoint is `Cola-DLM`. Use that exact identifier when checking documentation or a provider listing. [Source](https://huggingface.co/ByteDance-Seed/Cola-DLM) ### How can you access Cola-DLM? The catalog documents **Open weights** access for Cola-DLM. [Source](https://huggingface.co/ByteDance-Seed/Cola-DLM) As of the catalog check, this version was listed as available. [Source](https://huggingface.co/ByteDance-Seed/Cola-DLM) Open weights means the model weights are distributed for use under their stated terms. It does not establish a hosted API, an open-source license, or a hardware requirement. ### What can Cola-DLM receive and generate? Documented inputs: **text**. [Source](https://huggingface.co/ByteDance-Seed/Cola-DLM) Documented outputs: **text**. [Source](https://huggingface.co/ByteDance-Seed/Cola-DLM) Input support and output generation are separate capabilities. Reading an image does not establish image generation; a text output listing does not establish speech output. File sizes, formats and media limits depend on the documented implementation. ### What license is recorded for Cola-DLM? The recorded license label is **apache-2.0**. [Source](https://huggingface.co/ByteDance-Seed/Cola-DLM) Read the linked license and associated use terms for this exact version before redistribution, commercial deployment or fine-tuning. A label alone does not establish every permission, restriction or obligation, and it does not determine the terms of a separate hosted service. ### What distinguishes Cola-DLM? These details come from the linked publisher material for this model. Reported capabilities and training descriptions are publisher statements, not independent test results. **Model architecture**: Continuous latent diffusion language model combining a text VAE and block-causal Diffusion Transformer. [Source](https://huggingface.co/ByteDance-Seed/Cola-DLM/resolve/c1eafdd9cfd8064aeb917d569ef70a075b353eed/README.md) **Generation method**: Uses Flow Matching to transport a prior over continuous text latents. [Source](https://huggingface.co/ByteDance-Seed/Cola-DLM/resolve/c1eafdd9cfd8064aeb917d569ef70a075b353eed/README.md) **Released components**: The release contains an approximately 500M Text VAE with four encoder and four decoder blocks, plus an approximately 1.8B, 24-layer diffusion transformer. These are component sizes, not a combined checkpoint count. [Source](https://raw.githubusercontent.com/ByteDance-Seed/Cola-DLM/main/docs/model_card.md) **Training stage**: Research pretraining checkpoint without instruction tuning or RLHF; it is not a drop-in instruction-following chatbot. [Source](https://raw.githubusercontent.com/ByteDance-Seed/Cola-DLM/main/docs/model_card.md) **Short-prompt limitation**: Prompts shorter than one configured block (typically block\_size=16) disable classifier-free guidance on their first block and may produce weaker output. [Source](https://raw.githubusercontent.com/ByteDance-Seed/Cola-DLM/main/docs/model_card.md) ### What is documented about the Cola-DLM checkpoint? This evidence refers to `ByteDance-Seed/Cola-DLM` at recorded revision `c1eafdd9cfd8064aeb917d569ef70a075b353eed`, checked October 23, 2026. [Repository metadata](https://huggingface.co/api/models/ByteDance-Seed/Cola-DLM) Repository task tag: `text-generation`. A task tag identifies a distribution category; it does not measure performance on that task. [Repository metadata](https://huggingface.co/api/models/ByteDance-Seed/Cola-DLM) Listed library: `transformers`. Check that library's version and model support before loading the weights. [Repository metadata](https://huggingface.co/api/models/ByteDance-Seed/Cola-DLM) Repository language labels: en. These labels are not independent evaluations of language quality. [Repository metadata](https://huggingface.co/api/models/ByteDance-Seed/Cola-DLM) Repository access gate: `false`. This records the repository's gate setting when checked; licensing and use terms still apply. [Repository metadata](https://huggingface.co/api/models/ByteDance-Seed/Cola-DLM) Read the retrieved model card for setup and qualifications. [Model card](https://huggingface.co/ByteDance-Seed/Cola-DLM/resolve/c1eafdd9cfd8064aeb917d569ef70a075b353eed/README.md) **Catalog qualifications:** Official publisher weight listing; availability refers to this distribution, not a guaranteed hosted API. Base, SFT and research variants are identified by their exact published names. [Source 1](https://huggingface.co/ByteDance-Seed/Cola-DLM) · [Source 2](https://huggingface.co/api/models/ByteDance-Seed/Cola-DLM) · [Source 3](https://huggingface.co/ByteDance-Seed/Cola-DLM/resolve/c1eafdd9cfd8064aeb917d569ef70a075b353eed/README.md) **Evidence still missing here:** context or input capacity, knowledge and training-data cutoffs, and tested hardware requirements. Missing documentation is not evidence that a feature is unavailable. Headless Thrones has not independently benchmarked this model for this reference. Evaluate it on representative work before choosing it; these facts do not rank its accuracy, reliability, speed or suitability against another model.