Images
explorer
Alternatives to a picture you have, near or far.
<!-- Generated by apps/docs/scripts/gen-tools.mjs. Do not edit here: the words live in apps/web/src/lib/tools/explorer.tool.json (agent, hint) and packages/shared/src/tools/setting-hints.ts. --> On the rail as **explorer**. ## When to use it Turning one picture into alternatives without writing a prompt: variations that wander as far as you set, counter-proposals that each break one assumption, or four twins drawn again from scratch. The mode picks which. ## When to reach for something else It takes no prompt beyond notes pinned on the picture, so it cannot rewrite a scene or change one detail, and Antagonist is built to include a result worse than yours. A plain set of variations is the variations tool's. ## How to use it 1. Drop a picture onto the card's **Reference** slot, from the board or your computer. 2. Set **Imagination**, **Images** on the card. 3. Pick a mode on the card: **Tangents**, **Antagonist**, **Twin**. 4. Pick a model in the pill at the top of the card, or keep **GPT Image 2**. 5. Press **Run**. Each result lands on the board as its own item. ## On the card | Control | What it does | Choices | | - | - | - | | Reference | The picture the tool works from. Drop one from the board or your computer. | One picture | | Imagination | How far the results may wander from your picture. Low stays close to it; high lets each one go further. | 0 to 100 | | Images | How many pictures one run brings back. | 3, 4 or 6 | | 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 ### Tangents Up to 6 variations that wander from the picture, as far as Imagination says. Models in this mode: **GPT Image 2**, **GPT 2.5 Sunburst**. ### Antagonist Up to 6 counter-proposals, each breaking one assumption the picture makes. Models in this mode: **GPT Image 2**, **GPT 2.5 Sunburst**. ### Twin Four new pictures of the same thing, drawn again rather than copied. Models in this mode: **GPT 2.5 Sunburst**. ## 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) | The original GPT Image 2 grid engine, the v01 pick on the antagonist and tangents pills: one reference in, one call, a 2x4 sheet sliced into up to eight canvas images, with the tool supplying the axis prompt. | | [GPT 2.5 Sunburst](#gpt-25-sunburst) | The GPT Image 2.5 contact-sheet engine behind the grid tools: one reference in, one call out, a sheet sliced into up to eight canvas images, with the tool supplying the axis prompt. The v02 pick beside the GPT Image 2 original. | ### GPT Image 2 **Not the one for:** Exactly one reference image and no free prompt; the direction comes from the tool's slider. The v02 sibling runs GPT Image 2.5 at the same charge, so v01 is the pick for the older engine's rendering. Behind the settings button: | Setting | What it does | Choices | | - | - | - | | Quality | The overall finish level. Higher looks better but costs more and is slower. | Low, Medium, High | | Size | Output resolution. Large (2K) renders sharper cells at a higher credit cost. | Standard, Large, 4K | ### GPT 2.5 Sunburst **Not the one for:** Exactly one reference image and no free prompt: the direction comes from the tool's mode and sliders. The quality ladder stops at high, and every cell shares one sheet, so a bigger count means smaller pictures. Behind the settings button: | Setting | What it does | Choices | | - | - | - | | Quality | The overall finish level. Higher looks better but costs more and is slower. | Low, Medium, High | | Size | Output resolution. Large (2K) renders sharper cells at a higher credit cost. | Standard, Large, 4K |On the rail as explorer.
When to use it
Section titled “When to use it”Turning one picture into alternatives without writing a prompt: variations that wander as far as you set, counter-proposals that each break one assumption, or four twins drawn again from scratch. The mode picks which.
When to reach for something else
Section titled “When to reach for something else”It takes no prompt beyond notes pinned on the picture, so it cannot rewrite a scene or change one detail, and Antagonist is built to include a result worse than yours. A plain set of variations is the variations tool’s.
How to use it
Section titled “How to use it”- Drop a picture onto the card’s Reference slot, from the board or your computer.
- Set Imagination, Images on the card.
- Pick a mode on the card: Tangents, Antagonist, Twin.
- Pick a model in the pill at the top of the card, or keep GPT Image 2.
- Press Run. Each result lands on the board as its own item.
On the card
Section titled “On the card”| Control | What it does | Choices |
|---|---|---|
| Reference | The picture the tool works from. Drop one from the board or your computer. | One picture |
| Imagination | How far the results may wander from your picture. Low stays close to it; high lets each one go further. | 0 to 100 |
| Images | How many pictures one run brings back. | 3, 4 or 6 |
| 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 |
Tangents
Section titled “Tangents”Up to 6 variations that wander from the picture, as far as Imagination says.
Models in this mode: GPT Image 2, GPT 2.5 Sunburst.
Antagonist
Section titled “Antagonist”Up to 6 counter-proposals, each breaking one assumption the picture makes.
Models in this mode: GPT Image 2, GPT 2.5 Sunburst.
Four new pictures of the same thing, drawn again rather than copied.
Models in this mode: GPT 2.5 Sunburst.
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 | The original GPT Image 2 grid engine, the v01 pick on the antagonist and tangents pills: one reference in, one call, a 2x4 sheet sliced into up to eight canvas images, with the tool supplying the axis prompt. |
| GPT 2.5 Sunburst | The GPT Image 2.5 contact-sheet engine behind the grid tools: one reference in, one call out, a sheet sliced into up to eight canvas images, with the tool supplying the axis prompt. The v02 pick beside the GPT Image 2 original. |
GPT Image 2
Section titled “GPT Image 2”Not the one for: Exactly one reference image and no free prompt; the direction comes from the tool’s slider. The v02 sibling runs GPT Image 2.5 at the same charge, so v01 is the pick for the older engine’s rendering.
Behind the settings button:
| Setting | What it does | Choices |
|---|---|---|
| Quality | The overall finish level. Higher looks better but costs more and is slower. | Low, Medium, High |
| Size | Output resolution. Large (2K) renders sharper cells at a higher credit cost. | Standard, Large, 4K |
GPT 2.5 Sunburst
Section titled “GPT 2.5 Sunburst”Not the one for: Exactly one reference image and no free prompt: the direction comes from the tool’s mode and sliders. The quality ladder stops at high, and every cell shares one sheet, so a bigger count means smaller pictures.
Behind the settings button:
| Setting | What it does | Choices |
|---|---|---|
| Quality | The overall finish level. Higher looks better but costs more and is slower. | Low, Medium, High |
| Size | Output resolution. Large (2K) renders sharper cells at a higher credit cost. | Standard, Large, 4K |