Images
typography
A picture's text removed, set in a new font, or translated.
<!-- Generated by apps/docs/scripts/gen-tools.mjs. Do not edit here: the words live in apps/web/src/lib/tools/typography-remove.tool.json (agent, hint) and packages/shared/src/tools/setting-hints.ts. --> On the rail as **typography**. ## When to use it Clearing every label, caption, watermark and sign out of a picture, keeping the words exactly where they sit and setting them in a font of your choice, or writing them in another language in the same type and place. ## When to reach for something else Remove rebuilds only what a label was covering, so a large or central sign can leave a soft patch behind. Edit only changes lettering: it cannot move or resize the text block itself. Languages writes one language a run. ## How to use it 1. Drop a picture onto the card's **Image** slot, from the board or your computer. 2. Write what you want in **Except**. 3. Set **Language** on the card. 4. Pick a mode on the card: **Remove**, **Edit**, **Languages**. 5. Pick a model in the pill at the top of the card, or keep **GPT Image 2**. 6. Press **Run**. ## On the card | Control | What it does | Choices | | - | - | - | | Image | The picture the tool works from. Drop one from the board or your computer. | One picture | | Except | List any text the tool must keep, like a brand name or headline. Everything else gets removed. | Your words | | Language | The language the picture's words are written in. One per run. | 日本語, עברית, Deutsch, Français, Español, العربية, 中文, English, Italiano, Português, Nederlands, 한국어, Polski, Türkçe, Русский, Українська, हिन्दी, ไทย, Tiếng Việt, Bahasa Indonesia, Svenska, Dansk, Norsk, Suomi, Ελληνικά, Čeština | | Run | Starts the run. The number on it is what this run costs in credits, worked out before you press. | | | ×1 | Runs the card several times in a row with the same settings. | 1 to 99 | ## Modes ### Remove Strip every piece of text and rebuild the surfaces behind it Models in this mode: **GPT Image 2**. ### Edit Keep the text where it is and set it in a font you pick Models in this mode: **GPT Image 2**. ### Languages The picture's words written in the language you pick, in the same type and place, and read back Models in this mode: **Ideogram 4.5**. Settings this mode adds: | Setting | What it does | Choices | | - | - | - | | Quality | How hard the model works on the edit. The size of the picture never changes the price. | Draft, Low, Medium, High | | Read every word back | The result is read back and compared with the words asked for. One that reads differently is flagged, never landed as done. | On or off | | Full size | Only the words' areas are cut out and rewritten, so the picture keeps its size and every other pixel. | On or off | ## Models The pill at the top of the card picks the model. **GPT Image 2** is the one a new card starts on. | Model | Good for | | - | - | | [GPT Image 2](#gpt-image-2) | Two treatments for the text in a picture: Remove strips every label, watermark and sign and rebuilds what was behind them, with an Except field for words that must stay; Edit keeps the words in place and sets them in a font you pick. | | [Ideogram 4.5](#ideogram-45) | Inpaint's Tag mode: one change per tag, typed where it goes, and everything the words do not name comes back as it was. Also rewrites the text in a picture into another language in the same type, and takes up to four reference pictures. | ### GPT Image 2 **Not the one for:** It does not add new text to a blank surface; that is a labelling job. In Edit the words hold their position, size and colour and only the letterforms change, and it opens at high quality because small or dense text needs it. Behind the settings button: | Setting | What it does | Choices | | - | - | - | | Quality | The overall finish level. Higher looks better but costs more and is slower. | Medium, High | ### Ideogram 4.5 **Not the one for:** Not a restyle of the whole picture and not a new composition: it keeps the framing. The High tier is the expensive one, and an area can take at most three references. Behind the settings button: | Setting | What it does | Choices | | - | - | - | | Quality | The overall finish level. Higher looks better but costs more and is slower. | Medium, High |On the rail as typography.
When to use it
Section titled “When to use it”Clearing every label, caption, watermark and sign out of a picture, keeping the words exactly where they sit and setting them in a font of your choice, or writing them in another language in the same type and place.
When to reach for something else
Section titled “When to reach for something else”Remove rebuilds only what a label was covering, so a large or central sign can leave a soft patch behind. Edit only changes lettering: it cannot move or resize the text block itself. Languages writes one language a run.
How to use it
Section titled “How to use it”- Drop a picture onto the card’s Image slot, from the board or your computer.
- Write what you want in Except.
- Set Language on the card.
- Pick a mode on the card: Remove, Edit, Languages.
- Pick a model in the pill at the top of the card, or keep GPT Image 2.
- Press Run.
On the card
Section titled “On the card”| Control | What it does | Choices |
|---|---|---|
| Image | The picture the tool works from. Drop one from the board or your computer. | One picture |
| Except | List any text the tool must keep, like a brand name or headline. Everything else gets removed. | Your words |
| Language | The language the picture’s words are written in. One per run. | 日本語, עברית, Deutsch, Français, Español, العربية, 中文, English, Italiano, Português, Nederlands, 한국어, Polski, Türkçe, Русский, Українська, हिन्दी, ไทย, Tiếng Việt, Bahasa Indonesia, Svenska, Dansk, Norsk, Suomi, Ελληνικά, Čeština |
| Run | Starts the run. The number on it is what this run costs in credits, worked out before you press. | |
| ×1 | Runs the card several times in a row with the same settings. | 1 to 99 |
Remove
Section titled “Remove”Strip every piece of text and rebuild the surfaces behind it
Models in this mode: GPT Image 2.
Keep the text where it is and set it in a font you pick
Models in this mode: GPT Image 2.
Languages
Section titled “Languages”The picture’s words written in the language you pick, in the same type and place, and read back
Models in this mode: Ideogram 4.5.
Settings this mode adds:
| Setting | What it does | Choices |
|---|---|---|
| Quality | How hard the model works on the edit. The size of the picture never changes the price. | Draft, Low, Medium, High |
| Read every word back | The result is read back and compared with the words asked for. One that reads differently is flagged, never landed as done. | On or off |
| Full size | Only the words’ areas are cut out and rewritten, so the picture keeps its size and every other pixel. | On or off |
Models
Section titled “Models”The pill at the top of the card picks the model. GPT Image 2 is the one a new card starts on.
| Model | Good for |
|---|---|
| GPT Image 2 | Two treatments for the text in a picture: Remove strips every label, watermark and sign and rebuilds what was behind them, with an Except field for words that must stay; Edit keeps the words in place and sets them in a font you pick. |
| Ideogram 4.5 | Inpaint’s Tag mode: one change per tag, typed where it goes, and everything the words do not name comes back as it was. Also rewrites the text in a picture into another language in the same type, and takes up to four reference pictures. |
GPT Image 2
Section titled “GPT Image 2”Not the one for: It does not add new text to a blank surface; that is a labelling job. In Edit the words hold their position, size and colour and only the letterforms change, and it opens at high quality because small or dense text needs it.
Behind the settings button:
| Setting | What it does | Choices |
|---|---|---|
| Quality | The overall finish level. Higher looks better but costs more and is slower. | Medium, High |
Ideogram 4.5
Section titled “Ideogram 4.5”Not the one for: Not a restyle of the whole picture and not a new composition: it keeps the framing. The High tier is the expensive one, and an area can take at most three references.
Behind the settings button:
| Setting | What it does | Choices |
|---|---|---|
| Quality | The overall finish level. Higher looks better but costs more and is slower. | Medium, High |