Bluesky Thread

Nano Banana Pro

November 20, 2025 View original thread

Nano Banana Pro

A reasoning model image generator!

- multi turn
- google search grounding
- interleaved image & text

console.cloud.google.com/vertex-ai/pu...

console.cloud.google.com

You are using an unsupported browser. If you see unexpected behavior, upgrade to a supported browser. Learn more.

33 2

3 hours later

this is nuts. grounded infographics on demand

A photograph of the Golden Gate Bridge on a clear blue day, overlaid with white hand-drawn engineering sketches and labels. At the top left, handwritten text reads “GOLDEN GATE BRIDGE, 1937.” The bridge spans diagonally across the image, with its tall red-orange tower in the foreground and the second tower in the distance. The San Francisco skyline appears faintly on the left across the water.

Several annotated measurements appear around the bridge:
• Near the roadway: “DECK WIDTH 90 FT.”
• Below the suspended roadway: “CLEARANCE 220 FT.”
• Along the length between towers: “MAIN SPAN 4,200 FT.”
• Near the cables: “CABLE DIAMETER 36.375 IN.”
• Next to the far tower: “TOWER HEIGHT 746 FT.”
• Beside a circular cross-section illustration of bundled wires: “MAIN CABLE CROSS-SECTION, 27,572 WIRES.”
• Bottom right next to a cube-like sketch showing cables passing through a block: “ANCHORAGE SYSTEM.”

Additional chalk-like diagrams appear in the sky area above the bridge:
• A simple outline of a suspension bridge with arrows labeled “TENSION” along the cables and “COMPRESSION” at the towers.
• Curved cable shapes and schematic lines showing structural forces.

The annotations are styled like white blueprint notes drawn on top of a scenic landscape photograph.

9 1

teor takes

Teortaxes
（DeepSeek 推特... •3h
this is how actual multimodality looks. Images manipulated as effortlessly as text. This is what OpenAl wanted to do with 4o and, it seems, never did. Gemini's edge is increasing.

4 1

More like this