# CogVLM [Canonical model page](https://headlessthrones.com/models/z-ai-cogvlm) · [Model library](https://headlessthrones.com/models) Public catalog reference from Headless Thrones. Exact catalog ID: `z-ai-cogvlm`. 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 CogVLM 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/z-ai-cogvlm#model-answer-identity) · [Access](https://headlessthrones.com/models/z-ai-cogvlm#model-answer-access) · [Inputs & outputs](https://headlessthrones.com/models/z-ai-cogvlm#model-answer-modalities) · [License](https://headlessthrones.com/models/z-ai-cogvlm#model-answer-license) · [Model details](https://headlessthrones.com/models/z-ai-cogvlm#model-answer-research) · [Weight details](https://headlessthrones.com/models/z-ai-cogvlm#model-answer-repository) ### Which CogVLM does this page describe? CogVLM is a model record from **Z.ai** in the CogVLM family. [Source](https://huggingface.co/zai-org/CogVLM) The recorded model identifier or checkpoint is `CogVLM`. Use that exact identifier when checking documentation or a provider listing. [Source](https://huggingface.co/zai-org/CogVLM) ### How can you access CogVLM? The catalog documents **Open weights** access for CogVLM. [Source](https://huggingface.co/api/models/zai-org/CogVLM) As of the catalog check, this version was listed without confirmed general availability. [Source](https://huggingface.co/zai-org/CogVLM) 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 CogVLM receive and generate? Documented inputs: **text and image**. [Source](https://raw.githubusercontent.com/THUDM/CogVLM/main/README.md) Documented outputs: **text**. [Source](https://raw.githubusercontent.com/THUDM/CogVLM/main/README.md) 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 CogVLM? The recorded license label is **CogVLM Model License (model weights); Apache 2.0 (code)**. [Source](https://raw.githubusercontent.com/THUDM/CogVLM/main/README.md) 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 CogVLM? These details come from the linked publisher material for this model. Reported capabilities and training descriptions are publisher statements, not independent test results. **Repository contents**: This repository distributes several CogVLM checkpoints, including a 224-resolution pretrained base and grounding variants. [Source](https://raw.githubusercontent.com/THUDM/CogVLM/main/README.md) **Nominal component parameters**: CogVLM-17B consists of 10 billion visual parameters and 7 billion language parameters. [Source](https://raw.githubusercontent.com/THUDM/CogVLM/main/README.md) **Checkpoint distribution**: The public bundle contains cogvlm-base-224.zip, cogvlm-base-490.zip, cogvlm-chat-v1.1.zip, cogvlm-grounding-base.zip, cogvlm-grounding-generalist-v1.1.zip. Select the archive appropriate to the task. [Source](https://huggingface.co/api/models/zai-org/CogVLM) **SAT runtime**: The documented cogvlm-chat-v1.1 command uses the SAT runtime with FP16 and four-bit quantization; this is a named Chat deployment example, not a memory guarantee for every bundle variant. [Source](https://raw.githubusercontent.com/THUDM/CogVLM/main/README.md) ### What is documented about the CogVLM checkpoint? This evidence refers to `zai-org/CogVLM` at recorded revision `b9ccd2d71f6ee657b845e3e7ce76853f8b07b660`, checked October 23, 2026. [Repository metadata](https://huggingface.co/api/models/zai-org/CogVLM) 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/zai-org/CogVLM) **Catalog qualifications:** Official language-model listing; downloadable full weights not confirmed in this pass. [Source 1](https://huggingface.co/zai-org/CogVLM) · [Source 2](https://huggingface.co/api/models/zai-org/CogVLM) **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. Keep the exact version in view ## Other CogVLM models from Z.ai. These are separate records in the same publisher family. A shared family name does not establish equal capabilities or interchangeable licenses. ### [cogvlm-base-224-hf](https://headlessthrones.com/models/z-ai-cogvlm-base-224-hf) Open weights ### [cogvlm-base-490-hf](https://headlessthrones.com/models/z-ai-cogvlm-base-490-hf) Open weights ### [cogvlm-chat-hf](https://headlessthrones.com/models/z-ai-cogvlm-chat-hf) Open weights ### [cogvlm-grounding-base-hf](https://headlessthrones.com/models/z-ai-cogvlm-grounding-base-hf) Open weights ### [cogvlm-grounding-generalist-hf](https://headlessthrones.com/models/z-ai-cogvlm-grounding-generalist-hf) Open weights ### [cogvlm2-llama3-caption](https://headlessthrones.com/models/z-ai-cogvlm2-llama3-caption) Open weights