Models

Included Systemly models

On the included Systemly connection, the assistant offers five curated text models. The default is Balanced. A model can be changed beside the chat composer; your next message uses the new choice, including when you switch partway through a conversation.

Usage is based on provider costs and a shared monthly allowance. The model picker labels lower, moderate and higher usage options; it does not quote a fixed number of messages.

Which one should you choose?

Models on your own key

With a saved provider key, the picker shows that provider's suggested chat models. OpenRouter models are loaded from its live catalogue, with a small fallback list if the catalogue cannot be reached. A model ID entered in settings is also selectable even if it is not suggested.

Your provider decides which models the key can use and what it charges. Systemly records the calls and tokens but deducts no Systemly usage for them. See Bringing your own key for setup.

Image and video models

AI Studio's Images & video area is separate from the chat model picker, and it shows which model each render will use beside its Generate button.

• On the included AI usage, images use Gemini 2.5 Flash Image by default and video uses Veo 3.1 Lite, with Nano Banana 2, GPT Image 2, FLUX, Seedream, Veo 3.1 Fast, Kling 3.0, Seedance and Sora 2 Pro also available. Each render counts towards usage at what it actually cost, and a render that fails is not charged.
• On your own OpenRouter key, you can enter any image or video model OpenRouter offers, such as google/veo-3.1-fast. Usage is billed by OpenRouter.
• On your own OpenAI key, images use GPT Image 1 or DALL-E 3 and video uses Sora 2 or Sora 2 Pro. Usage is billed by OpenAI.

Video models accept different lengths and shapes. Systemly uses the length nearest to the one you pick that the model allows, and its closest supported shape. Video renders in the background, so you can leave the page and it appears in the library when it is done.

Making images and video for a workspace

• Ideas. Under the prompt box are ideas for this business, such as behind-the-scenes or before-and-after shots. Select one to fill in the prompt, then edit it. More ideas asks the AI for a fresh set written from the brand kit, which uses a small amount of the allowance.
• Brand kit switch. With Brand kit on, the business description, audience, visual style and brand colour are added to the prompt. For images, the logo and brand images are also sent to models that accept reference images (most OpenRouter image models do), which uses more of your allowance. Video models take no reference images, so for video the brand kit is added as words only.
• Viewing. Select an image or video in the library to see it full size. A render of several images shows them all; use the arrows or arrow keys to move between them.
• Using it. Download saves the file. Use in post opens the social composer with that image or video already attached, ready to caption, choose accounts and schedule. It needs a connected social account and permission to write posts.

Nothing generated is posted automatically. A person always chooses to post it.