Create images, video and audio with AI
Start with the result you want. Use a compatible model for an image, a clip, speech or another supported task, then inspect and refine it with saved versions.
Connect the account for the task
OpenRouter connects directly with its API key and does not require OpenCode. Direct text chat uses the conversation and explicit attachments, without local coding-agent tools. Your provider controls model access, credit and charges.
Give each input a clear role
Compatible models offer independent first frame, last frame, image, video and audio inputs. Check the chosen model before attaching files; its formats, parameters and lower limits still apply.
Keep each attempt understandable
Select a saved version to inspect it without changing the draft. Use inputs restores its prompt, references, account, model and settings. Generate version submits the next request. Extend video is available only for an eligible completed job and account.
Follow the detailed guides
- Connect OpenRouter directly
Connect your own OpenRouter account for direct text chat and specialized image, video and audio models.
- Generate images, video and audio
Choose a model, prepare independent frames and references, generate a result and inspect or download it.
- Versions, references and video extensions
Revisit saved generations, reuse their inputs and extend eligible videos with deliberate new submissions.