Google Nano Banana 2.1 Edit API Documentation

Google Nano Banana 2.1 Edit API Documentation

Playground

Google Nano Banana 2.1 Edit is the latest Nano Banana image editing model, with gains in visual design, subject consistency and prompt adherence, up to 14 reference images and 1K to 4K output. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.

Features

Nano Banana 2.1 Edit is Google’s latest Nano Banana AI image editing model. Google reports that it outperforms its previous image models across the board, with notable leaps in visual design and subject consistency, tighter prompt adherence and cleaner text rendering — so edits keep people and products recognizable.

Why Choose This?

• Google’s newest Nano Banana Upgraded over Nano Banana 2 in visual design, subject consistency and prompt adherence.

• Natural language editing Modify images using simple text instructions — the model understands context and relationships.

• Multi-image reference Upload up to 14 reference images for complex edits and compositions.

• Multi-resolution support Output in 1K, 2K, or 4K resolution based on your needs.

• Flexible aspect ratios Fourteen options: 1:1, 3:2, 2:3, 3:4, 4:3, 4:5, 5:4, 9:16, 16:9, 21:9, 1:4, 4:1, 1:8, 8:1.

• Format choice Export in PNG or JPEG format.

Parameters

How to Use

1. Upload reference images — add the images you want to edit (up to 14 images).
2. Write your prompt — describe the change, then what must stay the same.
3. Choose aspect ratio (optional) — select a preset or leave empty to follow the input.
4. Select resolution — choose 1K, 2K, or 4K based on your needs.
5. Choose output format — PNG or JPEG.
6. Run — submit and download your edited image.

Pricing

Example Costs

Best Use Cases

• Character Modification — Change clothing, setting, or style while keeping identity.
• Product Placement — Put a product into new scenes while keeping it accurate.
• Multi-reference Composition — Combine subjects, outfits and styles from several images.
• Text Editing — Modify or translate on-image text while keeping the design.
• Scene Adjustment — Change backgrounds, lighting, or environmental elements.

Pro Tips

• State the edit first, then what to keep: “replace the background with a beach at dusk, keep the person and lighting unchanged”.
• Put the main subject first and refer to other references by position (“the jacket from image 2”).
• Fewer reference images (1–3) give the most stable results for simple edits.

Notes

• Both images and prompt are required fields.
• Maximum reference images: 14.
• The first reference image is included; each additional reference image costs $0.003.
• Ensure your prompts comply with Google’s Safety Guidelines.

Related Models

• Nano Banana 2.1 Text-to-Image — Generate images from text prompts.
• Nano Banana 2 Edit — Previous generation with a 0.5K option and Fast variants.
• Nano Banana Pro Edit — Pro tier editing.

Authentication

For authentication details, please refer to the Authentication Guide.

API Endpoints

Submit Task & Query Result

set -euo pipefail

export WAVESPEED_API_KEY="your-api-key"

REQUEST_BODY=$(cat <<'JSON'
{
  "prompt": "A cinematic ocean wave at sunrise, highly detailed",
  "images": [
    "https://interactive-examples.mdn.mozilla.net/media/cc0-images/painted-hand-298-332.jpg"
  ],
  "aspect_ratio": "1:1",
  "resolution": "1k",
  "enable_web_search": false,
  "enable_image_search": false,
  "output_format": "png"
}
JSON
)

# 1. Submit the prediction.
SUBMIT_RESPONSE=$(curl --silent --show-error --fail-with-body \
  -X POST "https://api.wavespeed.ai/api/v3/google/nano-banana-2.1/edit" \
  -H "Authorization: Bearer ${WAVESPEED_API_KEY}" \
  -H "Content-Type: application/json" \
  -d "${REQUEST_BODY}")

TASK=$(printf '%s' "${SUBMIT_RESPONSE}" | jq 'if type == "object" and has("data") then .data else . end')
PREDICTION_ID=$(printf '%s' "${TASK}" | jq -r '.id // empty')
if [ -z "${PREDICTION_ID}" ]; then
  printf 'Submission response did not contain a prediction id
' >&2
  exit 1
fi
RESULT_URL="https://api.wavespeed.ai/api/v3/predictions/${PREDICTION_ID}/result"

# 2. Poll until the prediction finishes.
while true; do
  RESPONSE=$(curl --silent --show-error --fail-with-body \
    "${RESULT_URL}" \
    -H "Authorization: Bearer ${WAVESPEED_API_KEY}")
  RESULT=$(printf '%s' "${RESPONSE}" | jq 'if type == "object" and has("data") then .data else . end')
  STATUS=$(printf '%s' "${RESULT}" | jq -r '.status // empty')

  case "${STATUS}" in
    completed) printf '%s\n' "${RESULT}" | jq '.outputs'; break ;;
    failed|cancelled|timeout|deleted) printf '%s\n' "${RESULT}" | jq . >&2; exit 1 ;;
    *) sleep 2 ;;
  esac
done

Parameters

Task Submission Parameters

Request Parameters

Response Parameters

Result Request Parameters

Result Response Parameters