AI Studio for Deep Research, Images & Video

Clico Web is a browser based workspace for research, files, images, and video. You can ask a current question, bring in documents, create a finished visual, and keep the result in one conversation. The same chat can hold your request, source links, selected files, generated media, and follow up questions. You do not need a separate workflow for each task, but the tools available can still depend on your plan, the model you choose, and current service configuration.

Clico is one product with three focused surfaces. Web is the browser workspace, Desktop works across Mac apps, and the Chrome Extension works beside the page already open. See how Clico evolved.

Clico can create actual image files and finished MP4 videos. Image requests can return up to four completed outputs, while video requests begin as a render job and become playable when the stored asset is ready. A queued or rendering result is not a finished video. Clico Web shows the current status in the conversation, then displays the available image or video when generation completes. You can use the finished result inside the chat, download it, or include available media in a shared conversation.

Add the image you want to use and tell Clico what should stay consistent or what should change. Clico only sends reference images that are explicitly selected for the generation request, so an unrelated upload is not used automatically. The current image and video contracts accept up to five selected reference images. A reference can guide the subject, product, character, setting, composition, or visual style, but the exact result still depends on the request and the active generation path.

Yes. Image and video generation use saved task records. You can leave the conversation and reopen it later, and Clico Web will check the stored task status again when the generation identifiers are still available. This applies to image and video renders, not to arbitrary background work. A render may still fail, expire, or be canceled, so Clico only presents the asset as complete after the finished image or video has been returned and stored.

Deep Research Analyst uses the same file and web tools available in Clico Web. It can inspect relevant attachments, search the public web from more than one angle, open useful pages, compare evidence, and keep source links close to the claims they support. It is a specialist approach inside the current conversation, not a separate autonomous agent. If web search or page retrieval is unavailable, Clico should disclose that evidence gap instead of presenting an unsupported claim as verified.

Clico can extract text from PDF, DOCX, PPTX, TXT, MD, CSV, and XLSX files. Images, audio, and video can also be added when the selected chat model supports that input type. The current composer accepts up to ten attachments in one request, with a limit of eight MB for each file and twenty MB combined. If a model cannot read a selected file, choose another model that lists the required input before sending the request.

No. You can start with the recommended option and change models only when the task requires something different. The picker groups the current catalog by provider and shows whether a model can read images, PDF files, audio, or video. It also shows whether the model is available on your plan. Model names, access, and supported inputs can change, so use the current picker rather than treating an older model list as permanent.

Every plan includes Clico Web. Free currently includes 200 monthly credits, selected standard Web models, and standard routing. Pro currently includes 2,000 monthly credits, expanded Web model access, and enhanced routing. Max raises the monthly allowance to 5,000 credits for more frequent work. Annual plans are billed upfront while credits are issued monthly. Use the plan selector above and the current information in your account as the source for pricing, allowance, and access.

Yes. Recent Clico Web chats can be reopened through their saved conversation records. Earlier messages return with the chat, so you can ask for a summary, rewrite part of the answer, regenerate an assistant response, or continue the same task without explaining the full context again. Saved chats can also be deleted. When a reopened conversation contains an active image or video task, Clico Web can resume checking that render status.

A public share is a revocable snapshot, not a live collaborative chat. It can include messages, selected text, public source links, available uploaded or generated images and videos, and the names and kinds of other attachments. It does not expose the full contents of ordinary document attachments, unfinished media, or deleted assets. You can stop sharing later. When sharing is revoked, access to shared image and video files ends with it.

Clico does not add a Clico watermark or branding overlay to the generated image or video file it stores. The image becomes available after the generation returns an asset, and the finished MP4 can be played in Clico Web after rendering completes. This statement is about Clico's own storage and delivery path. It does not mean that a queued generation is finished, and it does not change any rights or usage terms that apply to the material in your request.