seedinteger
The seed for the random number generator.
Nano Banana 2 Lite Edit
google/nano-banana-2-lite/editNano Banana 2 Lite Edit is Google's low-latency, cost-efficient image editing model (Gemini 3.1 Flash Lite Image). Edit and combine supplied images with a text prompt; output is a 1K image.
Try the model
Playground
Input
The prompt for image editing.
PNG, JPEG, WebP, or GIF · 20 MiB maximum each
The URLs of the images to use for image-to-image generation or image editing.
Additional Settings More
Customize your input with more control.
The format of the generated image. Allowed values: jpeg, png.
Optional system instruction that steers the model's persona and output style across the request. Leave blank to omit; when provided, it is sent as the system instruction to Gemini (or as a system message on OpenAI-compatible providers).
When set, enables model thinking with the given level ('minimal' or 'high') and includes thoughts in the generation. Omit to disable. Allowed values: minimal, high.
The aspect ratio of the generated image. Leave empty to keep the aspect ratio of the input image. Allowed values: 21:9, 16:9, 3:2, 4:3, 5:4, 1:1, 4:5, 3:4, 2:3, 9:16.
The seed for the random number generator.
ResultExample
What would you like to do next?
Examples
Select an example to load its inputs and result into the playground.
- Open result ↗
Input
{ "images": [ "https://static.sandbase.ai/examples/17b5627b-8b12-4fc0-9e93-72f557c3fced/input/56496fc5-007b-4821-b507-b6b1b60e9408.png" ], "prompt": "make the background light blue" }
Human preference
Arena
Elo ratings from blind side-by-side votes, compared with other image and video models.
Rankings sourced from Artificial Analysis Arena
- Leaderboard
- Image Editing
- Rank
- #35 of 77
- Elo
- 1054 ±12
- Gap to #1
- −145
Image-editing arena: people pick the better of two edits of the same image and instruction, without seeing the model names.
- #30Qwen-Image-3.01068±13
- #31Qwen-Image-2.1-Turbo1064±14
- #32HunyuanImage 3.0 Instruct1063±13
- #33Seedream 5.0 Lite1062±13
- #34Qwen Image 2.0 Pro (2026-04-22)1058±17
- #35Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image)1054±12
- #36Wan 2.7 Pro1050±13
- #37DreamImage 2.01048±13
- #38Wan 2.6 Image1044±13
- #39Luma UNI 11043±13
Nano Banana 2 Lite Edit
Nano Banana 2 Lite Edit: fast, low-cost edits and multi-image composition at 1K with Gemini 3.1 Flash Lite Image.
Fields
| Field | Type | Default | Notes |
|---|---|---|---|
prompt | string, required | — | Up to 50,000 characters. |
images | image URL[], required | — | Images to edit or combine, Image 1, Image 2, … |
output_format | jpeg / png | png | Output file format. |
system_prompt | string | — | Optional standing instructions applied to every request. |
thinking_level | minimal / high | — | high reasons more about the prompt before drawing; minimal is faster. |
aspect_ratio | 21:9 / 16:9 / 3:2 / 4:3 / 5:4 / 1:1 / 4:5 / 3:4 / 2:3 / 9:16 | — | Omit to follow the first input image. |
seed | integer | random | Fix it to compare prompt changes. |
Billing
$0.0336 per 1K image, Google's list price.
| Request | Price |
|---|---|
| One image | $0.0336 |
Failed runs are not charged.
Prompting
- Keep prompts direct; use
thinking_level: highfor complex layouts.
Limits
- Output is 1K only.
Also on SandBase
Other Nano Banana endpoints: google/nano-banana-pro, google/nano-banana-pro/edit, google/nano-banana-2, google/nano-banana-2/edit, google/nano-banana-2-lite.
Start building
Send your first request
Submit this model through the SandBase Unified Run API.
# The quoted heredoc keeps Unicode and shell metacharacters unchanged.
result=$(curl --fail-with-body --silent \
-X POST "https://api.sandbase.ai/v1/run" \
-H "Authorization: Bearer $SANDBASE_API_KEY" \
-H "Content-Type: application/json" \
--data-binary @- <<'SANDBASE_JSON'
{
"model": "google/nano-banana-2-lite/edit",
"images": [
"https://static.sandbase.ai/examples/google/nano-banana-2-lite/edit/input_images_0.png",
"https://static.sandbase.ai/examples/google/nano-banana-2-lite/edit/input_images_1.png"
],
"prompt": "make a photo of the man driving the car down the california coastline",
"output_format": "png"
}
SANDBASE_JSON
)
run_id=$(printf '%s' "$result" | jq -r .id)
for attempt in $(seq 1 120); do
status=$(printf '%s' "$result" | jq -r .status)
case "$status" in completed|failed|timeout) break ;; esac
sleep 2
result=$(curl --fail-with-body --silent \
-H "Authorization: Bearer $SANDBASE_API_KEY" \
"https://api.sandbase.ai/v1/run/$run_id")
done
status=$(printf '%s' "$result" | jq -r .status)
[ "$status" = completed ] || { echo "Generation ended: $status" >&2; exit 1; }
printf '%s\n' "$result"2. Authentication
Create a SandBase API key and set it in your server environment. All requests use Bearer authentication.
export SANDBASE_API_KEY="YOUR_API_KEY"Keep the key on your server. Do not expose it in browser code or commit it to a repository.
3. Queue
Submit a request
POST to /v1/run with this model ID and its inputs. The request example above includes submission and polling.
Fetch request status and result
Use the returned run ID to poll until the status is completed, failed or timeout. Read output only after completed.
curl "https://api.sandbase.ai/v1/run/$RUN_ID" \
-H "Authorization: Bearer $SANDBASE_API_KEY"4. Files
Hosted files
Pass a publicly accessible file URL in the model’s image, video or audio field. Supported formats and limits are shown beside each Playground input.
Uploading images
Use the upload or replace control on supported image fields in the Playground. Video URL fields accept a hosted driving video; they do not imply video-upload support.
5. Schema
Input
imagesarrayrequired
The URLs of the images to use for image-to-image generation or image editing.
promptstringrequired
The prompt for image editing.
aspect_ratiostring
The aspect ratio of the generated image. Leave empty to keep the aspect ratio of the input image.
Possible values: "21:9", "16:9", "3:2", "4:3", "5:4", "1:1", "4:5", "3:4", "2:3", "9:16"
output_formatstring
The format of the generated image.
Default: "png"
Possible values: "jpeg", "png"
system_promptstring
Optional system instruction that steers the model's persona and output style across the request. Leave blank to omit; when provided, it is sent as the system instruction to Gemini (or as a system message on OpenAI-compatible providers).
Default: ""
thinking_levelstring
When set, enables model thinking with the given level ('minimal' or 'high') and includes thoughts in the generation. Omit to disable.
Possible values: "minimal", "high"
{
"model": "google/nano-banana-2-lite/edit",
"images": [
"https://static.sandbase.ai/examples/google/nano-banana-2-lite/edit/input_images_0.png",
"https://static.sandbase.ai/examples/google/nano-banana-2-lite/edit/input_images_1.png"
],
"prompt": "make a photo of the man driving the car down the california coastline",
"output_format": "png"
}Output
A completed run exposes its result in output. The following is the model’s saved example, not a new generation.
[
{
"source": "Self organization completed prediction",
"source_url": "https://www.sandbase.ai",
"url": "https://static.sandbase.ai/examples/17b5627b-8b12-4fc0-9e93-72f557c3fced/output/7a8ec137-6183-4b5c-9d57-6680f2b29c30.jpg"
}
]