Rivya History Guide

Use this history guide when you need to find previous chats, generation results, task records, saved context, downloads, or the exact place to continue a project.

History in Rivya is not there so the product can say it “has history.”

It is there because AI work becomes much more useful once you can return to the exact point where it stopped making sense to keep everything in your head.

The First Question History Helps You Answer

When users open history, they are usually not asking for an archive.

They are asking one of two things:

or:

That is why Rivya splits history into two different views instead of forcing everything into one mixed feed.

Why There Are Two History Surfaces

The current history split is:

• /history/generations
• /history/chats

That mirrors the product’s real split:

• image, video, and audio are async tasks
• chat is saved as sessions and messages

If those were mixed together, the useful next action would be much harder to find.

Generation History Is for Results and Next Actions

Generation history is where image, video, and audio work stays usable after the moment passes.

Each generation record can show things like:

• the model
• a prompt preview
• task status
• credit cost
• refund state
• the main result preview

The important part is not just what it shows. It is what it lets you do next.

Depending on the result, generation history can help you:

• reopen the matching workspace
• download the output
• take an image result into video
• open a result in chat for follow-up

So generation history is not just where outputs go to sit. It is where outputs become the next step.

Chat History Is for Re-entering Context

Chat history solves a different problem.

It stores sessions rather than tasks, which means it is built for reopening thinking, not only reopening output.

A chat item can currently show:

• session title
• chat model
• tool badge, when the session started from a tool
• updated time
• last-message preview

That is enough to answer a very practical question:

When the answer is yes, chat history brings you back into /studio/chat/session/[sessionId], not into a generic blank workspace.

History Is Best Used Right After You Lose the Thread

The most realistic time to use history is not “once a week for reporting.”

It is right after one of these moments:

• you left a generation page before the task finished
• you know you already made something similar and do not want to start over
• you remember the conversation, but not the exact model or session
• you want to continue from an old output instead of reprompting from scratch

That is where history stops being an archive and starts acting like project memory.

History vs Dashboard

Use dashboard when you want the quick operational picture.

Use history when you already know you need one of these:

• an old generation
• an old chat session
• a more specific re-entry point than the dashboard gives you

Dashboard tells you what is recent. History helps you reopen it deliberately.

History vs Notifications

The clean distinction is:

• history answers “what did I make or discuss?”
• notifications answer “what important event happened?”

These overlap, but they are not interchangeable.

If a generation failed, notifications may tell you that it failed. History is where you look when you want to inspect that task as part of the larger body of work.

A Good Way to Use History

For most real users, the most useful history pattern is:

1. open history when you know the work already exists
2. choose between chats and generations first
3. reopen the exact item instead of starting over
4. only leave history once you have the next workspace open

That sounds obvious, but it is the difference between history as a shelf and history as workflow memory.

Read Next

• Task Lifecycle
• Image Workflows in Rivya
• Video Workflows in Rivya
• Audio Workflows in Rivya
• Notifications Center
• Dashboard in Rivya
• Chat Workflows in Rivya

History Review Checklist

When you need to find, reuse, or explain past work, check:

• Check the correct history kind: chat, image, video, audio, or tool-related chat.
• Confirm whether the task completed, failed, expired, or never started upstream.
• Use notifications and task status when history alone does not explain what happened.
• Save useful outputs before starting variants or switching models.
• Keep billing and credit questions tied to the transaction or task that created them.

Recheck Before Assuming Output Is Lost

Recheck if the user changed accounts, filters, history type, model path, or browser session. Many missing result cases are actually wrong-surface or wrong-status cases.