AI Chat

Genspark · Official agreement

Read the agreement

54 of 54 cards

Use & Processing · Context

AI Chat is a text chat with one model per conversation and its own 5-hour limit

Genspark describes AI Chat as a chat for text: you ask and it answers in the conversation, cites what it used when search is on, reads files you attach, and drafts or rewrites text you paste in. Each chat has one model behind it, picked before the first message and kept for the length of the chat. AI Chat has its own model picker and its own 5-hour limit, which are not shared with Super Agent or Free Quota. Super Agent is a separate product.

Supporting quotations

You ask, and it answers in the conversation. When search is on it cites what it used. It reads the files you attach, and it drafts or rewrites text you paste in.
It has its own model picker and its own 5-hour limit. Neither is shared with Super Agent or with Free Quota.
You pick the model before the first message and keep it for the length of the chat.
It has its own model picker and its own 5-hour limit.
The two don't share a picker or a limit.

Qualifications & exceptions

  • Citations appear only when search is on.
  • The picker and the 5-hour limit belong to AI Chat and are not shared with Super Agent or Free Quota.
Also annotated in the same passage (1)

Use & Processing · Company practice

AI Chat reaches the web, attached files and Genspark's product documentation, not Hub files or past chats

AI Chat can reach three things: the web through Search Web, the files you attach, and Genspark's own product documentation, which it can read whether or not search is on. It does not read your Hub files, and it has no memory of your earlier chats.

Supporting quotations

AI Chat can reach three things: the web through Search Web, the files you attach, and Genspark's own product documentation, which it can read whether or not search is on.
It doesn't read your Hub files, and it has no memory of your earlier chats.
Genspark's own product documentation is always available to AI Chat, whether or not Search Web is on.

Qualifications & exceptions

  • Web access is through the Search Web switch.
  • Genspark's product documentation is available whether or not search is on.
Also annotated in the same passage (1)

Use & Processing · Restriction or exception

The model is chosen before the first message and locks after it

Quick start lists a Genspark account and a question as what you need. You open the picker before typing; a fast model is selected by default, and a new chat opens on the model you last picked in this browser if it is still offered. Once you send the first message the picker's model list is locked and reads "You need to start a new chat to switch models" with a New Chat button. To change model you open a new chat and pick before typing; the old chat keeps its model and its history. Whether a model uses credits depends on your plan.

Supporting quotations

Before you start: a Genspark account and a question. Pick the model before you type, because the picker locks after the first message.
Pick the model first, because it locks after your first message
A fast model is selected by default, and a new chat opens on the model you last picked in this browser if it's still offered.
Whether a model uses credits depends on your plan
From then on the picker's model list is locked
Search Web and Personalization below the list still work.
The old chat keeps its model and its history.

Qualifications & exceptions

  • Search Web and Personalization below the model list still work after the lock.
  • The default or last-picked model applies only if it is still offered.
  • Whether a model uses credits depends on the plan.

Use & Processing · User right or choice

Search Web can be turned on later but not off once it has run; Personalization needs a profile

Search Web and Personalization are switches in the picker dropdown. Search Web can be turned on later in the chat. Once search has run it cannot be turned off, except with Mixture-of-Agents. Personalization appears once you have filled in a profile.

Supporting quotations

Search Web can be turned on later in the chat too. Once search has run, it can't be turned off, except with Mixture-of-Agents. Personalization appears once you've filled in a profile.

Qualifications & exceptions

  • Mixture-of-Agents is the exception to the rule that search cannot be turned off.
  • Personalization shows only after a profile exists.

Charges & Financial Obligations · Context

The 5-hour limit and the model credit rules apply to every kind of job

The article says two sections apply to every job in AI Chat: the AI Chat 5-hour limit and which models cost credits. They are explained in the sections that follow.

Supporting quotations

Two sections right after Quick start apply to every job: the AI Chat 5-hour limit and which models cost credits.

Use & Processing · Context

The Genspark home-page composer can start Super Agent rather than AI Chat

The home-page composer is another way into AI Chat, but it can start Super Agent, so the article tells you to check which agent the box names before typing a long prompt into it. Super Agent has its own picker and limit.

Supporting quotations

The Genspark home page composer is another way in, but it can start Super Agent, so check which agent the box names before you type a long prompt into it.

Qualifications & exceptions

  • The article does not state how a prompt typed into the wrong agent is treated.

Account & Service Control · Company practice

Team and Enterprise seats keep every AI Chat model at no credit cost; limits differ by seat type

Team and Enterprise seats keep every AI Chat model at no credit cost, whatever date the seat started. Team seats have the 5-hour limit, the same size as Plus, and stop at it. Enterprise seats, and Team organizations billed by invoice, have no AI Chat limit. The picker shows Recommended and then models grouped by maker. An Enterprise admin can limit which model vendors appear; the FAQ adds that vendor restrictions apply by vendor, not by individual model.

Supporting quotations

Team and Enterprise seats keep every AI Chat model at no credit cost, whatever date the seat started.
Enterprise seats, and Team organizations billed by invoice, have no AI Chat limit.
An Enterprise admin can limit which model vendors appear.
You can't restrict individual models.

Qualifications & exceptions

  • Team seats stop at the limit; Enterprise seats and invoiced Team organizations have no AI Chat limit.
  • Vendor restrictions work by vendor only; individual models cannot be restricted.

Cross-references

Also annotated in the same passage (1)

Charges & Financial Obligations · Company practice

Credits come from one wallet shared by every Genspark product

Credits come from one wallet that every Genspark product draws on, so a credit spent in AI Chat is a credit Super Agent cannot use.

Supporting quotations

Credits come from one wallet that every Genspark product draws on, so a credit spent in AI Chat is a credit Super Agent can't use.
Also annotated in the same passage (1)

Account & Service Control · Context

The model lineup changes often and what the picker shows varies by plan, region, organization and rollout

There are no Standard or Ultra tiers in AI Chat. The picker has a search box and lists models by maker. On Plus and Pro from September 18, 2026 a No credit cost list of core models comes first, four models first with More opening the rest and flagship models below; on every other plan a short Recommended list comes first. Each row shows its cost as a multiplier. The article says the lineup changes often, so the picker is the list to trust, and the models you see can vary by plan, region, organization settings and gradual rollouts. Mixture-of-Agents is listed under Genspark near the bottom.

Supporting quotations

There are no Standard or Ultra tiers in AI Chat.
On every other plan a short Recommended list comes first.
The lineup changes often, so the picker is the list to trust. The models you see can vary by plan, region, organization settings and gradual rollouts.
Every row shows its cost as a multiplier such as "2×"
If your picker opens on a No credit cost section, you're on Plus or Pro from September 18, 2026.

Qualifications & exceptions

  • The layout differs between Plus and Pro from September 18, 2026 and every other plan.
  • Availability can vary by plan, region, organization settings and gradual rollouts.

Not stated here

  • The article does not list the models currently offered or their costs.
Also annotated in the same passage (4)

Use & Processing · Company practice

A chat keeps its model to the end unless the model is retired, then it moves to the replacement

The model is chosen before the first message. After that the model list is dimmed, and opening it reads "You need to start a new chat to switch models" with a New Chat button that opens a new chat in a new tab. A chat keeps its model to the end, unless that model is retired; then the chat moves to its replacement.

Supporting quotations

The model is chosen before the first message.
A chat keeps its model to the end, unless that model is retired. Then the chat moves to its replacement.

Qualifications & exceptions

  • The move to a replacement happens only when the model is retired.

Not stated here

  • The article does not say here how much notice is given before a model is retired.

Charges & Financial Obligations · Company practice

Mixture-of-Agents runs several models, hides its mix, and its credit cost can depend on that mix

Mixture-of-Agents is listed under Genspark in the picker. Several models answer the same question and an aggregator writes one reply from their answers. It is slower than a single model. The mix is not shown in the picker, differs with Search Web on or off, and changes over time. On Plus and Pro from September 18, 2026 it counts as a core model; on a free account every model in the mix uses credits. The article advises picking a single model if you need to be sure of the cost. Elsewhere it says the whole message bills credits if the mix includes a flagship model.

Supporting quotations

Several models answer the same question and an aggregator writes one reply from their answers.
It's slower than a single model, because more than one model runs.
The mix isn't shown in the picker. It differs with Search Web on or off and it changes over time. On Plus and Pro from September 18, 2026, it counts as a core model. On a free account every model in the mix uses credits. If you need to be sure of the cost, pick a single model.
The pick is Mixture-of-Agents and the mix includes a flagship model, so the whole message bills.

Qualifications & exceptions

  • Counts as a core model only on Plus and Pro from September 18, 2026.
  • On a free account every model in the mix uses credits.
  • The mix differs with Search Web on or off and changes over time.

Not stated here

  • Because the mix is not shown, the article does not say how a user can tell whether it includes a flagship model.

Cross-references

Also annotated in the same passage (2)

Use & Processing · Company practice

With Search Web on, replies link sources, but how closely models follow this varies and you should check them

Search Web is a switch in the picker dropdown, described as "Retrieve from Internet". With it on, replies put a [Source] link after the sentence or paragraph that relies on a search result, on every plan. How closely a model follows this varies, and Mixture-of-Agents replies do not carry the links. The article advises opening two or three of the links to check that each source says what the reply says.

Supporting quotations

Search Web is a switch in the picker dropdown, described as "Retrieve from Internet". Turn it on before the first message or at any point later in the chat.
With Search Web on, replies put a [Source] link after the sentence or paragraph that relies on a search result, on every plan. How closely a model follows this varies, and Mixture-of-Agents replies don't carry the links.
Open two or three of the links and check that each source says what the reply says.

Qualifications & exceptions

  • How closely a model follows the source-link behaviour varies.
  • Mixture-of-Agents replies do not carry the links.

Cross-references

Use & Processing · Restriction or exception

Once search has run in a chat it cannot be turned off, except with Mixture-of-Agents; some models cannot search

Once search has run in a chat you cannot turn it off; the switch then reads "Search cannot be disabled after using search tools", and you start a New Chat for a chat without it. Mixture-of-Agents is the exception: its search switch stays free. If a model cannot search, the Search Web switch is grayed out and stays off, and you choose another model in a new chat.

Supporting quotations

Once search has run in a chat, you can't turn it off.
Mixture-of-Agents is the exception: its search switch stays free.
If a model can't search, the Search Web switch is grayed out and stays off. Choose another model in a new chat.

Qualifications & exceptions

  • Applies once search has run in that chat.
  • Mixture-of-Agents keeps its search switch free.
  • Models that cannot search keep the switch grayed out and off.

Use & Processing · Restriction or exception

AI Chat accepts listed file types up to 100 MB each and 80 images per conversation

AI Chat reads what you attach and accepts PDF, EPUB, CSV, MD, TXT, HTML, JSON, XML, YAML, Word, Excel, PowerPoint, code files and images. The limit is 100 MB per file and 80 images per conversation. Google Docs, Sheets and Slides cannot be attached as they are; you download the file as Word, PDF or Excel first and attach that.

Supporting quotations

AI Chat accepts PDF, EPUB, CSV, MD, TXT, HTML, JSON, XML, YAML, Word, Excel, PowerPoint, code files and images. The limit is 100 MB per file, and 80 images per conversation.
Google Docs, Sheets and Slides can't be attached as they are. Download the file as Word, PDF or Excel first, then attach that.

Qualifications & exceptions

  • Limits are 100 MB per file and 80 images per conversation.
  • Google Docs, Sheets and Slides must be downloaded as Word, PDF or Excel first.

Not stated here

  • The article does not say how long attached files are kept or who can access them; it does not address storage of attachments.

Use & Processing · Company practice

Large text files are read in parts, and models that cannot read images reject them

Large text files are read in parts, on demand. Most models read images. With a model that cannot, the file picker does not offer images, and pasting one shows "This model does not support images".

Supporting quotations

Large text files are read in parts, on demand.
Most models read images. With a model that can't, the file picker doesn't offer images, and pasting one shows "This model does not support images".

Qualifications & exceptions

  • Image support depends on the chosen model.

Use & Processing · User right or choice

Personalization is on by default and adds your profile to each reply until you switch it off for a chat

Personalization is on by default. It adds what you have written in Settings → Personalization to each reply: your nickname, occupation, profile and chosen traits, plus your custom instructions on paid plans. You can turn it off for a chat with the Personalization switch in the model picker, under Search Web; the switch reads "Apply your profile preferences" and appears once you have filled in your nickname, occupation, traits or profile. Each chat remembers its own setting and a new chat starts with the setting you used last in this browser. Editing the profile applies to every chat. It is a profile you set, not a memory of past chats; AI Chat has no memory system and does not read your Hub files.

Supporting quotations

Personalization is on by default. It adds what you've written in Settings → Personalization to each reply: your nickname, occupation, profile and chosen traits, plus your custom instructions on paid plans.
Turn it off for a chat with the Personalization switch in the model picker, under Search Web.
Each chat remembers its own setting, and a new chat starts with the setting you used last in this browser.
You edit the profile in Settings → Personalization, and it applies to every chat.
Personalization is a profile you set, not a memory of past chats. AI Chat has no memory system, and it doesn't read your Hub files.

Qualifications & exceptions

  • Custom instructions are included on paid plans only.
  • The off switch appears only once a profile field is filled in.
  • The setting is per chat; a new chat starts with the last setting used in this browser.

Not stated here

  • The article does not say whether profile details are shared with the model vendors serving a reply beyond saying they are added to each reply.

Use & Processing · Company practice

Genspark's product documentation is a separate tool that AI Chat can always use

Genspark's own product documentation is a separate tool from search. It is always available to AI Chat, whether or not Search Web is on, which is how AI Chat answers questions about Genspark products.

Supporting quotations

Genspark's own product documentation is a separate tool
That's how it answers questions about Genspark products.
Also annotated in the same passage (1)

Use & Processing · Company practice

Long chats are compacted automatically and can stop with a context error that requires a new chat

AI Chat compacts older history automatically as a chat gets long. When it does, the chat shows "For better performance, previous chat history has been compacted." Older history is compacted to make room, so the chat keeps going. If a chat still overflows, the reply stops with "Context Length Exceeded, Please open a new session" and a Create a new session button, which is the point to start a new chat. There is no ring or compress control in AI Chat; those belong to Super Agent.

Supporting quotations

AI Chat compacts older history automatically as a chat gets long.
Older history is compacted to make room, so the chat keeps going.
If a chat still overflows, the reply stops with "Context Length Exceeded, Please open a new session" and a Create a new session button.
There's no ring or compress control in AI Chat: the context ring and the compress button belong to Super Agent, and they don't appear here.

Qualifications & exceptions

  • Compaction happens automatically; AI Chat has no manual compress control.
  • A chat that still overflows stops and requires a new chat.

Not stated here

  • The article does not say which parts of older history are dropped or summarized when a chat is compacted.
Also annotated in the same passage (1)

Use & Processing · Restriction or exception

AI Chat answers and writes; files, computer work and pictures go to other Genspark products by copy and paste

AI Chat answers and writes. When the job needs a file, a computer or a picture, another product takes it, and there is no button from AI Chat to any of them, so the handoff is a copy and a paste. The article's table points to Super Agent for a report, deck, sheet or document built from a brief; AI Docs for a document you create and edit with AI; AI Sheets for a spreadsheet; Custom Agent for an agent created from one prompt; Claw for a personal assistant across your apps; Realtime Voice for a spoken conversation with Super Agent; Speakly for dictation; and AI Image or Super Agent for pictures.

Supporting quotations

AI Chat answers and writes. When the job needs a file, a computer or a picture, another product takes it. There's no button from AI Chat to any of them, so the handoff is a copy and a paste.
A report, deck, sheet or document built from a brief
A document you create and edit with AI
A spreadsheet built from data and a prompt
An agent you create from one prompt
A personal assistant that works across your apps
A spoken conversation with Super Agent
Dictation in any text box
A picture made or edited

Qualifications & exceptions

  • The handoff is manual: no button links AI Chat to the other products.

Not stated here

  • The linked product pages are not supplied, so their terms and costs are not described here.

Charges & Financial Obligations · Context

Plan, credit, seat-pricing and troubleshooting detail sits on other Genspark pages not supplied here

The article points to the Membership Plans and Credits Guide pages for plan and credit detail, to the Team & Enterprise Plans page for seat pricing, and to the Troubleshooting page for an error, an outage or a model that vanished. Those pages are linked but not part of this document.

Supporting quotations

Plan and credit detail
An error, an outage or a model that vanished
has seat pricing

Not stated here

  • Plan prices, credit amounts per model and seat pricing are not stated in this article.
Also annotated in the same passage (2)

Use & Processing · Context

The troubleshooting table lists the usual cause of common problems, including no OCR on scanned files

The table gives what each symptom "usually" is: hitting the limit with credits left means the 5-hour window is spent and the plan decides what happens next; credits taken on a "No credit cost" model have three causes named in the FAQ; the chat not loading, sending or finishing is a timeout on Genspark's side, not your account; a compressed chat that lost the thread is automatic compaction, which is normal in a long chat; a context error means the chat ran out of room; a looping or wrong reply is usually fixed with a new chat or another model; on a computer the web mic opens the Speakly site; and a scanned file comes back empty because AI Chat does not run OCR on attached files.

Supporting quotations

The 5-hour window is spent, and the plan decides what happens next
Three causes, named in the FAQ entry
A timeout on Genspark's side, not your account
Automatic compaction, which is normal in a long chat
The chat ran out of room
A new chat or another model is the usual fix
On a computer, the web mic opens the Speakly site
A scan is a picture of text, and AI Chat doesn't run OCR on attached files

Qualifications & exceptions

  • The causes are described as what each symptom usually is, not as exhaustive.

Use & Processing · Context

A model can disappear because it is retired or withdrawn, and models cannot be switched inside a chat

If a model you used has disappeared, the article says it is being retired and the picker hides it in the days before, or it was withdrawn. You cannot switch models inside a chat because the picker locks after the first message; use another model on the same topic in a new chat.

Supporting quotations

It's being retired, and the picker hides it in the days before, or it was withdrawn
The picker locks after the first message

Qualifications & exceptions

  • Retirement hides the model in the picker in the days before it is retired; a withdrawn model is a separate case.

Not stated here

  • The Troubleshooting page that explains retired models is linked but not supplied.

Charges & Financial Obligations · Context

The 5-hour window and the credit balance are separate counters

The window and your credit balance are separate counters, so you can hit the limit with credits left. The message in the composer shows which one stopped the chat. Credits come from one wallet that every Genspark product draws on.

Supporting quotations

The window and your credit balance are separate counters. The message in the composer shows which one stopped the chat.
Credits come from one wallet that every Genspark product draws on

Qualifications & exceptions

  • What happens when the window is spent depends on the plan.
Also annotated in the same passage (1)

Charges & Financial Obligations · Company practice

A Regenerate response that has to run the request again uses credits like any message

If something fails on Genspark's side, the reply shows "There was an error generating a response." with a Regenerate response button. Regenerate response first tries to recover the reply; if it has to run the request again, that run uses credits like any message. If it keeps failing, the article says to wait a few minutes and reload, and points to its site-won't-load (524) and app-slow pages.

Supporting quotations

Regenerate response first tries to recover the reply. If it has to run the request again, that run uses credits like any message.
If something fails on Genspark's side, the reply shows "There was an error generating a response." with a Regenerate response button.

Qualifications & exceptions

  • Credits are used only if Regenerate response has to run the request again.

Not stated here

  • The article does not say whether a failed reply that is not recovered is credited back.

Use & Processing · Procedure

After a context error the fix is a new chat that does not inherit the old history

When a chat stops with a context error, the article says to start a new chat. To avoid this, ask for a summary before the chat fills up. The new chat does not inherit the old one's history. A new topic deserves a new chat, and there is no ring or compress control in AI Chat, since those belong to Super Agent.

Supporting quotations

Start a new chat. To avoid this next time, ask for a summary before the chat fills up.
The new chat doesn't inherit the old one's history.
A new topic deserves a new chat. There's no ring or compress control in AI Chat: the context ring and the compress button belong to Super Agent, and they don't appear here.
That's the point to start a new chat.

Qualifications & exceptions

  • The new chat starts without the old history.
Also annotated in the same passage (1)

Use & Processing · Context

AI Chat answers in the conversation; Super Agent runs multi-step jobs and delivers results, often as files

AI Chat answers in the conversation. Super Agent takes on a multi-step job, uses tools such as web research, files and your connected apps to carry it out, and delivers the result, often as files. The article advises using AI Chat to think the ask through and Super Agent to execute it. The two do not share a picker or a limit. AI Chat is a chat for text that answers questions, researches with web search, reads attached files, and writes or edits text.

Supporting quotations

It's a chat for text. It answers questions, researches with web search, reads the files you attach, and writes or edits text.
AI Chat answers in the conversation. Super Agent takes on a multi-step job, uses tools such as web research, files and your connected apps to carry it out, and delivers the result, often as files.
The two don't share a picker or a limit.

Qualifications & exceptions

  • The two products do not share a picker or a limit.
Also annotated in the same passage (1)

Charges & Financial Obligations · Context

The picker layout depends on the plan and each row shows what it costs

The FAQ says to open the picker before the first message. On Plus and Pro from September 18, 2026 it has two groups, No credit cost and Flagship models; on every other plan it shows Recommended and then models grouped by maker. Each row shows what it costs. Which messages bill credits depends on the model and on the plan, as the credits-by-plan section explains.

Supporting quotations

On Plus and Pro from September 18, 2026, it has two groups, No credit cost and Flagship models. On every other plan it shows Recommended and then models grouped by maker. Each row shows what it costs.
The picker has a search box and lists models by maker.
Which one applies depends on the model and on the plan.
Every row shows its cost as a multiplier such as "2×"

Qualifications & exceptions

  • The two-group layout applies to Plus and Pro from September 18, 2026 only.
Also annotated in the same passage (4)

Use & Processing · Restriction or exception

AI Chat produces text only and cannot generate images, though it reads images you attach

AI Chat cannot generate images. It produces text only, though it reads images you attach. For a picture, the article points to AI Image or Super Agent.

Supporting quotations

No. It produces text only, though it reads images you attach. For a picture, use AI Image or Super Agent.

Data Access & Collection · Company practice

Voice input opens the Speakly site on a computer and records into text on phones and tablets

On a computer, the mic button in the message box opens the Speakly site in a new tab. Speakly is Genspark's dictation app, and once installed it types what you say into any text box, including AI Chat. On a phone or tablet browser, the mic records your voice and turns it into text in the message box. The article advises reading the text before sending it so a misheard word does not go through. Live voice conversation is a Super Agent feature.

Supporting quotations

On a computer, the mic button in the message box opens the Speakly site in a new tab. Speakly is Genspark's dictation app, and once it's installed it types what you say into any text box, including AI Chat.
On a phone or tablet browser, the mic records your voice and turns it into text in the message box.
Live voice conversation is a Super Agent feature.

Qualifications & exceptions

  • Behaviour differs between a computer and a phone or tablet browser.
  • On a computer the Speakly app must be installed for dictation to work.

Not stated here

  • The article does not say where recorded voice is processed or whether audio is kept.

Retention & Deletion · Company practice

AI Chat has no memory system; chats are saved as projects and only your profile applies to every chat

AI Chat does not remember past conversations: it has no memory system and does not read your Hub files. Your chats are saved as projects, so ending a conversation does not delete it. Your Personalization profile is the one thing applied to every chat, and it is a profile you set, not a memory of what you discussed.

Supporting quotations

No. It has no memory system, and it doesn't read your Hub files. Your chats are saved as projects, so ending a conversation doesn't delete it.
Your Personalization profile is the one thing applied to every chat, and that's a profile you set, not a memory of what you discussed.

Qualifications & exceptions

  • The Personalization profile is the only item applied across chats.

Charges & Financial Obligations · Context

Credits can be taken on a "No credit cost" model for three reasons on newer Plus and Pro plans

For Plus and Pro plans from September 18, 2026, whose picker shows "No credit cost", credits can be taken for three reasons: the model picked is a flagship model, which always bills credits; the pick is Mixture-of-Agents and the mix includes a flagship model, so the whole message bills; or the 5-hour window was already spent and core models were running on credits, in which case the core header reads "Limit reached". On a free account every message uses credits.

Supporting quotations

This applies to Plus and Pro plans from September 18, 2026, the plans whose picker shows "No credit cost". Three things cause it.
The model you picked is a flagship model, which always bills credits.
The pick is Mixture-of-Agents and the mix includes a flagship model, so the whole message bills.
Or the 5-hour window was already spent, and core models were running on credits. The core section header then reads "Limit reached". On a free account every message uses credits.
every AI Chat model is free, flagship included

Qualifications & exceptions

  • Applies to Plus and Pro plans from September 18, 2026.
  • Plans from before that date have every model free until December 31, 2026 under the other section.
Also annotated in the same passage (4)

Retention & Deletion · User right or choice

An AI Data Retention switch sits under Settings → Account; Team and Enterprise plans are opted out

The AI Data Retention switch is under Settings → Account. On a Team or Enterprise plan the setting reads "Because you are on a Team or Enterprise plan, you are opted out of AI Data Retention."

Supporting quotations

The AI Data Retention switch is under Settings → Account. On a Team or Enterprise plan, the setting reads "Because you are on a Team or Enterprise plan, you are opted out of AI Data Retention."

Qualifications & exceptions

  • On Team and Enterprise plans the setting shows an opt-out rather than a choice.

Not stated here

  • The article does not say what AI Data Retention covers, what its default is for other plans, or what changes when the switch is flipped.

Retention & Deletion · Company practice

Chats are kept until you delete them; deleting removes a chat from your list and Delete User removes all data

Chats are kept until you delete them and do not expire. Deleting a chat removes it from your chat list, and it cannot be opened again. To remove all of your data, use Delete User under Settings → Account. How and when deleted data is erased is described in the Privacy Policy, which the article links.

Supporting quotations

Chats are kept until you delete them, and they don't expire. Deleting a chat removes it from your chat list, and it can't be opened again. To remove all of your data, use Delete User under Settings → Account.
How and when deleted data is erased is described in the

Qualifications & exceptions

  • Deleting a chat removes it from the chat list and it cannot be opened again.
  • Organizations that keep project history disable deletion for their members.

Not stated here

  • How and when deleted data is erased is left to the Privacy Policy, which is not supplied in this document.

Retention & Deletion · Restriction or exception

An organization can keep project history, which turns off deleting chats for its members

When an organization keeps project history, deletion is turned off. In the chat's "..." menu, Delete is replaced by "Your organization keeps project history. Deleting projects and conversations is disabled." An admin turns this on in Team admin › Policies › Project history › Keep project and conversation history. While it is on, files can still be edited or removed, and projects can still be removed from workspaces.

Supporting quotations

Your organization keeps project history, which turns deletion off.
An admin turns this on in Team admin › Policies › Project history › Keep project and conversation history. While it's on, files can still be edited or removed, and projects can still be removed from workspaces.

Qualifications & exceptions

  • Set by an admin; applies while the setting is on.
  • Files can still be edited or removed and projects removed from workspaces.
Also annotated in the same passage (1)

Account & Service Control · Company practice

Team admins can see credit usage by member and product and export usage logs as CSV

Admins see credit usage in Team admin › Usage analytics, where Member Stats breaks usage down by member and by product, and Usage Logs can be exported as CSV. AI Chat messages that a seat covers use no credits, so AI Chat may show little or no credit usage there. Seat pricing is on the Team & Enterprise Plans page, and the billing rule is in the Team and Enterprise seats section.

Supporting quotations

Admins see credit usage in Team admin › Usage analytics, where Member Stats breaks usage down by member and by product, and Usage Logs can be exported as CSV.
AI Chat messages that a seat covers use no credits, so AI Chat may show little or no credit usage there.

Qualifications & exceptions

  • Seat-covered AI Chat messages use no credits, so usage shown there may be little or none.

Not stated here

  • The article does not say what message content, if any, admins can see in the usage logs.
Also annotated in the same passage (2)

Account & Service Control · Company practice

Admins can restrict AI model vendors, and a chat on a blocked vendor continues on an allowed model

An admin can control which models people use only by vendor; individual models cannot be restricted. In Team admin › Connectors the AI Model Vendor Restrictions card has a Restrict AI Model Vendors switch. When it is on, members can only use models from the selected vendors and AI Chat's picker hides the rest. If a chat was started on a model whose vendor is no longer allowed, the next reply comes from an allowed model instead. The card appears once vendor restrictions are enabled for the organization, which requires contacting Genspark.

Supporting quotations

Yes, by vendor. You can't restrict individual models.
When it's on, members can only use models from the vendors you select, and AI Chat's picker hides the rest. If a chat was started on a model whose vendor is no longer allowed, the next reply comes from an allowed model instead.
The card appears once vendor restrictions are enabled for your organization, so contact Genspark to turn them on.

Qualifications & exceptions

  • Restrictions apply by vendor only.
  • The card appears only after Genspark enables vendor restrictions for the organization.
Also annotated in the same passage (1)

Charges & Financial Obligations · Procedure

Recipe 1: the picker layout tells you your plan, and flagship models cost credits or are locked

The recipe says to look at how the picker is laid out: if it opens on a No credit cost section followed by Flagship models, you are on Plus or Pro from September 18, 2026; if it opens on Recommended and then maker groups, the core and flagship split does not apply. It suggests a core model for a first pass and a flagship model in a new chat for a second opinion. On the newer Plus and Pro plans flagship models use credits, and on some free accounts they show locked with an upgrade prompt.

Supporting quotations

If it opens on a No credit cost section followed by Flagship models, you're on Plus or Pro from September 18, 2026. If it opens on Recommended and then maker groups, that split doesn't apply to you
On the newer Plus and Pro plans flagship models use credits, and on some free accounts they show locked with an upgrade prompt.
If your picker opens on a No credit cost section, you're on Plus or Pro from September 18, 2026.
The model you picked is a flagship model, which always bills credits.
flagship rows also carry a Credits tag
Each row shows what it costs.

Qualifications & exceptions

  • The core and flagship split applies only to Plus and Pro from September 18, 2026.
  • Locked flagship models with an upgrade prompt appear on some free accounts only.
Also annotated in the same passage (6)

Use & Processing · Restriction or exception

Recipe 3: Google files must be downloaded first, and scanned files come back empty because there is no OCR

The file recipe says a Google Doc, Sheet or Slide has to be downloaded first, as Word, PDF or Excel. It also says a scanned file comes back empty because AI Chat does not run OCR on attached files; the suggested workarounds are to attach the pages as images to a model that reads images, or to export a text-based PDF.

Supporting quotations

A Google Doc, Sheet or Slide has to be downloaded first, as Word, PDF or Excel.
a scanned file comes back empty because AI Chat doesn't run OCR on attached files. Attach the pages as images to a model that reads images, or export a text-based PDF.

Qualifications & exceptions

  • Workaround via images works only with a model that reads images.

Use & Processing · Procedure

Recipe 4: source links appear on their own, but the recipe also tells you to open and check them

The search recipe says to switch Search Web on before the first message, read the [Source] link beside each claim (Search Web adds them on its own), open two or three links and check that each source says what the reply says, and leave search on for the rest of the chat. It says you get an answer with a source for each claim, while the web search section says how closely a model follows the source-link behaviour varies. If the Search Web switch is grayed out, the model cannot search and you start a new chat with another model.

Supporting quotations

Open the picker and switch Search Web on before the first message.
Read the [Source] link beside each claim. Search Web adds them on its own.
Open two or three of the links and check that each source says what the reply says.
Leave search on for the rest of the chat.
You get: an answer with a source for each claim.
if the Search Web switch is grayed out, that model can't search. Start a new chat with another model.

Qualifications & exceptions

  • The recipe says to turn search on before the first message; the web search section says it can also be turned on later.
  • A grayed-out switch means that model cannot search.

Not stated here

  • The article does not reconcile "an answer with a source for each claim" with the statement that how closely a model follows source links varies.

Cross-references

Use & Processing · Procedure

Editing a message with the pencil regenerates the reply from that point and replaces the original text

The rewrite recipe says that to change one part you use the pencil on your message, edit it and send again, and the reply is regenerated from there. It also says the original text stays in your message above the rewrite unless you edited that message with the pencil.

Supporting quotations

To change one part, use the pencil on your message, edit it and send again. The reply is regenerated from there.
with your original text still in your message above it, unless you edited that message with the pencil

Qualifications & exceptions

  • The original text is kept only if the message was not edited with the pencil.

Not stated here

  • The article does not say whether a regenerated reply after an edit uses credits or counts toward the 5-hour window.

Charges & Financial Obligations · Company practice

When a model is retired, the chat moves to the replacement and messages are billed at the replacement rate

A chat that uses a retiring model shows a notice above the message box: "{model} will be retired on {date}. Switch to {replacement} to avoid interruption." It appears only in chats that use the retiring model, can be dismissed, and comes back in the model's last 7 days. You can start a New Chat on the replacement or keep going in the old chat; after the retirement date its next reply comes from the replacement automatically. Either way, messages are billed at the replacement's rate, which may differ from what the old model cost. The old chat's history stays in your chat list and remains readable. If the replacement is not in your picker, it may not be offered on your plan, in your region, or by your organization's model policy.

Supporting quotations

It appears only in chats that use the retiring model. You can dismiss it, and it comes back in the model's last 7 days.
After the retirement date, its next reply comes from the replacement automatically. Either way, messages are billed at the replacement's rate, which may differ from what the old model cost.
Keep the old chat for reference. Its history stays in your chat list.
the old conversation still readable and, after the retirement date, able to continue on the replacement
if the replacement isn't in your picker, it may not be offered on your plan, in your region, or by your organization's model policy

Qualifications & exceptions

  • The notice appears only in chats that use the retiring model.
  • Billing follows the replacement model's rate after the switch.
  • The replacement may be unavailable on some plans, regions or organization policies.

Not stated here

  • The article does not state how far ahead of the retirement date the notice first appears, only that it comes back in the last 7 days.
  • The Troubleshooting page that explains why models retire is linked but not supplied.

Content & IP Rights · Procedure

Save to AI Drive stores a reply as a Markdown file and may ask free-plan users to upgrade

The Save an answer recipe is for a reply you want outside the chat. Copy under the reply is for anything you will paste elsewhere. In the "..." menu, Save to AI Drive keeps the reply as a Markdown file in the Genspark › AI Chat folder of your AI Drive; on a free plan this may ask you to upgrade first. A saved reply is named after the chat with the date and time added, and the chat is unchanged. If Save to AI Drive asks you to upgrade, the article says your plan does not include saving files out of a chat, so use Copy instead.

Supporting quotations

For: a reply you'll want next week, outside the chat.
Use Copy under the reply for anything you'll paste elsewhere.
Choose Save to AI Drive to keep the reply as a Markdown file in the Genspark › AI Chat folder of your AI Drive. On a free plan, this may ask you to upgrade first.
A saved reply is named after the chat, with the date and time added.
If Save to AI Drive asks you to upgrade, your plan doesn't include saving files out of a chat, so use Copy instead.

Qualifications & exceptions

  • On a free plan Save to AI Drive may ask you to upgrade first.

Not stated here

  • The article does not say how long saved files are kept or how AI Drive storage is limited.
Also annotated in the same passage (1)

Sharing & Disclosure · Procedure

Save to Notion puts a reply in a Notion page and first starts a Notion connection if none exists

Choose Save to Notion to put the reply in a page. Until Notion is connected, that row reads Connect Notion and starts the connection first; after connecting you save again.

Supporting quotations

Choose Save to Notion to put it in a page. Until Notion is connected, that row reads Connect Notion and starts the connection first.
If the Notion row reads Connect Notion, connect Notion first and save again.

Qualifications & exceptions

  • A Notion connection must exist before saving.

Not stated here

  • The article does not describe what access the Notion connection grants Genspark.
Also annotated in the same passage (1)

Account & Service Control · Company practice

Chats are saved as projects; you can pin up to 50, delete only chats you own, and add chats to Hub

Every AI Chat conversation is saved in the sidebar chat list as a project. You can pin chats, with a limit of 50 ("You can only pin up to 50 projects. Please unpin some first."), rename or bookmark them, or delete them; Delete appears only on chats you own. Add to Hub adds the chat to your Hub. If you cannot delete a chat, your organization may keep project history, and the chat menu then shows "Your organization keeps project history. Deleting projects and conversations is disabled."

Supporting quotations

Every AI Chat conversation is saved there as a project.
The limit is 50: "You can only pin up to 50 projects. Please unpin some first."
Delete appears only on chats you own.
Add to Hub adds the chat to your
your organization may keep project history. The chat's menu then shows "Your organization keeps project history. Deleting projects and conversations is disabled." where Delete would be.
Your organization keeps project history, which turns deletion off.

Qualifications & exceptions

  • Delete appears only on chats you own.
  • Deletion is disabled where an organization keeps project history.

Not stated here

  • The Hub page is linked but not supplied, so who can see chats added to the Hub is not stated here.
Also annotated in the same passage (1)

Data Access & Collection · Procedure

Dictation: on a computer it runs through the Speakly app; on phones and tablets the browser needs microphone access

For dictation on a computer you need the Speakly app; on a phone or tablet you need a browser that allows microphone access. On a computer clicking the mic opens the Speakly site in a new tab, where you install Speakly, and Speakly types your dictation into the message box. On a phone or tablet recording starts in the message box and the voice input turns it into text when you finish. The article advises reading the text before sending it. On a computer the mic always opens the Speakly site and does not record in the page, because desktop dictation runs through the Speakly app.

Supporting quotations

on a computer, the Speakly app. On a phone or tablet, a browser that allows microphone access.
On a computer, this opens the Speakly site in a new tab. Install Speakly from there, then come back to the chat. On a phone or tablet, recording starts in the message box.
On a phone or tablet, the voice input turns it into text when you finish.
on a computer the mic always opens the Speakly site and doesn't record in the page, because desktop dictation runs through the Speakly app.

Qualifications & exceptions

  • Computer and phone or tablet behave differently.
  • A phone or tablet browser must allow microphone access.

Not stated here

  • The article does not say where voice is processed or whether audio is retained; the Speakly page is linked but not supplied.

Account & Service Control · Restriction or exception

AI Chat has its own fixed 5-hour usage window per user, separate from Free Quota

AI Chat has a usage limit of its own: a fixed 5-hour window, per user, that starts with your first message that counts toward it. New chats do not reset it. When the window ends, the counter starts again with your next message. The window covers AI Chat only and is not part of Free Quota, which has its own weekly limits; AI Chat draws on neither, and other agents keep working while AI Chat is limited. Free accounts, Enterprise seats and Team organizations billed by invoice do not have the window; Team seats have it, the same size as Plus, and stop at it.

Supporting quotations

AI Chat has a usage limit of its own: a fixed 5-hour window, per user, that starts with your first message that counts toward it.
Free accounts, Enterprise seats and Team organizations billed by invoice don't have it. New chats don't reset it. When the window ends, the counter starts again with your next message.
The window covers AI Chat only, and it isn't part of
Free Quota has its own weekly limits, and AI Chat draws on neither. Other agents keep working while AI Chat is limited.
Team seats have the 5-hour limit, the same size as Plus, and stop at it.

Qualifications & exceptions

  • Free accounts, Enterprise seats and Team organizations billed by invoice do not have the window.
  • Team seats have the window, the same size as Plus, and stop at it.
  • The window starts with the first message that counts toward it.

Not stated here

  • The article does not state the number of messages or amount of model spend a Plus or Pro window holds; it says only that Pro holds more than Plus.

Cross-references

Charges & Financial Obligations · Context

AI Chat shows one of four limit states, from free core models to a hard stop, depending on plan

Depending on the plan, AI Chat tells you where you stand in one of four ways. (1) "Core models at no credit cost": the window has room and core models are free. (2) "Core models now use credits" with the reset time: the window is spent, so core models now bill credits. (3) A message saying you have used up the 5-hour limit for core models and run out of credits, offering credits, an upgrade or waiting for the reset: the window is spent and the credit balance is empty; Pro accounts see the same line without "upgrade your plan". (4) A plan that stops at the limit shows the reset time in a Session Limit Reached dialog and as a reply.

Supporting quotations

Depending on your plan, AI Chat tells you where you stand in one of four ways. On Plus and Pro from September 18, 2026, a strip above the composer shows the first three while a core model is selected, and you can close it with ✕.
The window has room, and core models are free.
The window is spent, so core models now bill credits.
The window is spent and the credit balance is empty.
Pro accounts see the same line without "upgrade your plan".
which carries the plan-level wording

Qualifications & exceptions

  • On Plus and Pro from September 18, 2026, a strip above the composer shows the first three states while a core model is selected, and it can be closed with the X.
  • The fourth state appears only on plans that stop at the limit.

Not stated here

  • The "Membership Plans" page, which the article says carries the plan-level wording of the last message, is linked but not supplied here.
Also annotated in the same passage (1)

Account & Service Control · Restriction or exception

On plans that stop at the limit, nothing runs until the reset and buying credits does not lift it

On plans that stop at the limit, AI Chat shows a reply in the chat and a Session Limit Reached dialog saying the limit is reached and when it resets. Your plan stops at the limit, so nothing runs until the reset, and buying credits does not lift it. The article says which plans block and which keep running in its section on which models cost credits, by plan.

Supporting quotations

On plans that stop at the limit, the fourth comes as a reply in the chat and a Session Limit Reached dialog.
Your plan stops at the limit, so nothing runs until the reset.
Wait for the time shown. Buying credits doesn't lift it.
Which plans block and which keep running is in Which models cost credits, by plan.

Qualifications & exceptions

  • Applies only to plans that stop at the limit.
Also annotated in the same passage (1)

Account & Service Control · Company practice

Moving from Plus to Pro resets the window once; a downgrade keeps what is left, capped at the new limit

Moving from Plus to Pro during a window resets it once, so Pro's larger window starts clean. A downgrade keeps what you have left, capped at the new plan's limit. The article also says Pro has a higher limit than Plus, so its window holds more model spend.

Supporting quotations

Moving from Plus to Pro during a window resets it once, so Pro's larger window starts clean. A downgrade keeps what you have left, capped at the new plan's limit.
Pro has a higher limit than Plus, so its window holds more model spend.

Qualifications & exceptions

  • The reset applies when moving from Plus to Pro during a window.
  • On a downgrade what remains is capped at the new plan's limit.

Charges & Financial Obligations · Company practice

Each AI Chat message is free inside the window or billed in credits, depending on model and plan

Every message is billed one of two ways: free inside the window, or on credits. Which one applies depends on the model and on the plan. If your picker opens on a No credit cost section you are on Plus or Pro from September 18, 2026; free accounts, Plus and Pro from before that date, and Team and Enterprise seats see Recommended followed by models grouped by maker. AI Chat messages that a Team or Enterprise seat covers use no credits.

Supporting quotations

Every message is billed one of two ways: free inside the window, or on credits. Which one applies depends on the model and on the plan.
If your picker opens on a No credit cost section, you're on Plus or Pro from September 18, 2026.
On every other plan it shows Recommended and then models grouped by maker.
AI Chat messages that a seat covers use no credits

Qualifications & exceptions

  • The picker layout identifies which plan rules apply.
  • Seat-covered messages on Team and Enterprise use no credits.

Not stated here

  • Plan-level and credit detail is on the Membership Plans and Credits Guide pages, which are linked but not supplied.
Also annotated in the same passage (5)

Charges & Financial Obligations · Company practice

On newer Plus and Pro plans, core models are free inside the window and flagship models always use credits

On Plus and Pro from September 18, 2026 the picker splits models in two. Core models cost no credits while the window has room. Flagship models always use credits and do not use up the window. Other plans do not have the split. The core section is headed "No credit cost" while the window has room and "Limit reached" when it is spent; it shows four models first and More opens the rest, with flagship models grouped by maker below. Every row shows its cost as a multiplier such as "2×", and flagship rows also carry a Credits tag. A Mixture-of-Agents message counts as core unless the mix includes a flagship model, and then the whole message bills credits.

Supporting quotations

On Plus and Pro from September 18, 2026, the picker splits models in two. Core models cost no credits while the window has room. Flagship models always use credits, and they don't use up the window. Other plans don't have this split.
On those plans the core section is headed "No credit cost" while the window has room and "Limit reached" when it's spent.
Every row shows its cost as a multiplier such as "2×", and flagship rows also carry a Credits tag.
A Mixture-of-Agents message counts as core unless the mix includes a flagship model, and then the whole message bills credits.
Or the 5-hour window was already spent, and core models were running on credits.

Qualifications & exceptions

  • Applies to Plus and Pro from September 18, 2026 only.
  • Once the window is spent, core models run on credits and the core header reads "Limit reached".
  • A Mixture-of-Agents message bills credits if its mix includes a flagship model.

Cross-references

Also annotated in the same passage (4)

Charges & Financial Obligations · Company practice

Free accounts use credits on every message, with 100 credits a day and no 5-hour window

On a free account every message uses credits, and a free account gets 100 credits a day for it. There is no 5-hour limit on a free account. The picker shows Recommended and then models grouped by maker, and pricier models cost more credits per message. Some models are for members only: a free account sees them but cannot run them, and on some accounts flagship models are locked behind Unlock top models.

Supporting quotations

Every message uses credits, and a free account gets 100 credits a day for it. There's no 5-hour limit on a free account.
Some models are for members only: a free account sees them but can't run them, and on some accounts flagship models are locked behind Unlock top models.
pricier models cost more credits per message

Qualifications & exceptions

  • Flagship models are locked behind Unlock top models on some accounts only.
  • Members-only models are visible but cannot be run on a free account.

Not stated here

  • The article does not say how the 100 daily credits relate to Free Quota, which it says AI Chat does not draw on.

Charges & Financial Obligations · Company practice

Plus and Pro from September 18, 2026 get free core models inside the window, then credits

If your subscription started on or after September 18, 2026, or you came back after a lapse of more than 14 days, core models are free inside the 5-hour window. When the window is spent, core models keep working on credits, so the chat keeps going until your credits run out. Then the composer says so, and buying credits lets you continue. Flagship models always use credits.

Supporting quotations

If your subscription started on or after September 18, 2026, or you came back after a lapse of more than 14 days, core models are free inside the 5-hour window.
When the window is spent, core models keep working on credits, so the chat keeps going until your credits run out. Then the composer says so, and buying credits lets you continue. Flagship models always use credits.

Qualifications & exceptions

  • Applies to subscriptions started on or after September 18, 2026, or after a lapse of more than 14 days.
  • Flagship models always use credits.

Cross-references

Charges & Financial Obligations · Company practice

Plus and Pro from before September 18, 2026 keep every model free until December 31, 2026

If your subscription started before September 18, 2026 and stayed active, every AI Chat model is free, flagship included, and the window blocks the chat when it is spent. A lapse of up to 14 days still counts as continuous. The article says this benefit runs until December 31, 2026 and that from January 1, 2027 "the rules above apply".

Supporting quotations

If your subscription started before September 18, 2026 and stayed active, every AI Chat model is free, flagship included, and the window blocks the chat when it's spent. A lapse of up to 14 days still counts as continuous.
That benefit runs until December 31, 2026. From January 1, 2027 the rules above apply.

Qualifications & exceptions

  • Applies to subscriptions that started before September 18, 2026 and stayed active; a lapse of up to 14 days counts as continuous.
  • The benefit runs until December 31, 2026; the article announces a change from January 1, 2027.

Not stated here

  • The article says "the rules above apply" from January 1, 2027 without naming which section; the preceding section covers Plus and Pro from September 18, 2026.
  • The article states no effective date or version for itself.

Cross-references

Also annotated in the same passage (1)