Meshy V6 Lite Multi Image To 3d API Documentation

Meshy V6 Lite Multi Image To 3d API Documentation

Playground

Meshy 6 Lite Multi-Image-to-3D turns 1 to 4 views of one object into a textured 3D model with Meshy’s lightweight model. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.

Features

Meshy 6 Lite Multi-Image-to-3D generates a 3D model from 1–4 views of the same object. Configure textures, topology, polygon count, and optional character rigging or animation in a single workflow.

Why Choose This?

• Multi-view 3D generation — Build a 3D model from up to 4 views of the same object.
• Optional texturing — Generate either an untextured mesh or a textured model.
• Texture guidance — Guide the final appearance with a text prompt or reference image.
• Mesh controls — Choose triangle or quad topology and set a target polygon count.
• Character preparation — Generate A-pose or T-pose characters for rigging workflows.
• Optional rigging and animation — Add a character skeleton and preset animation when needed.

Parameters

How to Use

1. Upload reference views — Provide 1–4 images of the same object from different angles.
2. Choose texture settings — Generate a textured model or disable texturing for an untextured mesh.
3. Configure topology optional — Choose triangle or quad topology and adjust the target polygon count.
4. Choose pose optional — Use an A-pose or T-pose when preparing a humanoid character for rigging.
5. Enable rigging or animation optional — Add a skeleton or apply a preset animation.
6. Submit — Generate and retrieve the available 3D outputs.

Pricing

Base Price

Character Options

Animation includes $0.20 for rigging and $0.12 for one animation.

Rigging is charged only once when both rigging and animation are enabled.

Example Costs

Best Use Cases

• Product reconstruction — Create 3D assets from several product views.
• Game assets — Generate textured or optimized models for real-time workflows.
• Character generation — Create humanoid characters prepared in A-pose or T-pose.
• Rigged characters — Generate characters ready for downstream animation.
• Animated assets — Apply a preset animation directly to generated humanoid models.
• 3D prototyping — Quickly create models from several visual references.

Pro Tips

• Use images of the same object with consistent appearance and proportions.
• Show the object from clearly different angles for better reconstruction.
• Use texture_prompt when you want explicit control over the material or surface appearance.
• Use texture_image when a visual reference is more useful than a written description.
• Keep should_remesh=true when you need control over topology and polygon count.
• Use pose_mode for humanoid characters that will be rigged or animated.
• Use a clearly structured humanoid character with visible limbs for better rigging results.

Notes

• images supports 1–4 views of the same object.
• should_texture defaults to true.
• When should_texture=false, texture-related options are not applied.
• If both texture_prompt and texture_image are supplied, texture_prompt takes priority.
• target_polycount supports 100–300000 and applies when should_remesh=true.
• The final polygon count may differ from the requested target.
• should_remesh defaults to true for Meshy 6 Lite.
• symmetry_mode is retained for compatibility but no longer affects the output.
• auto_size estimates the object’s real-world height and rescales the model accordingly.
• Rigging and animation are intended for textured humanoid characters with clearly defined limbs and body structure.
• Rigging supports models up to 380000 triangles; each quad counts as two triangles.
• enable_animation=true includes the rigging step and applies one preset action using animation_action_id.

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'
{
  "images": [
    "https://interactive-examples.mdn.mozilla.net/media/cc0-images/painted-hand-298-332.jpg"
  ],
  "should_texture": true,
  "enable_pbr": false,
  "topology": "triangle",
  "target_polycount": 30000,
  "should_remesh": true,
  "symmetry_mode": "auto",
  "pose_mode": "a-pose",
  "auto_size": false,
  "enable_rigging": false,
  "rigging_height_meters": 1.7,
  "enable_animation": false,
  "animation_action_id": 92
}
JSON
)

# 1. Submit the prediction once. Never retry this request automatically:
# a retry after an unclear failure can create a second, billed task.
CHECK='The task may already exist. Check History or the predictions list before submitting again.'
if ! SUBMIT_RESPONSE=$(curl --silent --show-error --fail-with-body \
  --connect-timeout 10 --max-time 60 \
  -X POST "https://api.wavespeed.ai/api/v3/meshy/v6-lite/multi-image-to-3d" \
  -H "Authorization: Bearer ${WAVESPEED_API_KEY}" \
  -H "Content-Type: application/json" \
  -d "${REQUEST_BODY}"); then
  printf 'Submit failed. %s\n' "${CHECK}" >&2
  exit 1
fi

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 "${PREDICTION_ID}" ]; then
  printf 'Submission response did not contain a prediction id. %s\n' "${CHECK}" >&2
  exit 1
fi
RESULT_URL="https://api.wavespeed.ai/api/v3/predictions/${PREDICTION_ID}/result"

# 2. Poll until the prediction reaches a final status.
BODY_FILE=$(mktemp)
trap 'rm -f "${BODY_FILE}"' EXIT
while true; do
  # Transient errors (429, 5xx) are retried, honoring Retry-After.
  HTTP_CODE=$(curl --silent --show-error --fail-with-body \
    --connect-timeout 10 --max-time 60 --retry 5 --retry-delay 2 \
    --output "${BODY_FILE}" --write-out '%{http_code}' \
    "${RESULT_URL}" \
    -H "Authorization: Bearer ${WAVESPEED_API_KEY}" || true)
  case "${HTTP_CODE}" in
    2??) ;;
    000|429|5??)  # dropped connection or temporary error: keep polling
      printf 'Task %s: HTTP %s, polling again in 10s\n' "${PREDICTION_ID}" "${HTTP_CODE}" >&2
      sleep 10; continue ;;
    *) printf 'Task %s: HTTP %s\n' "${PREDICTION_ID}" "${HTTP_CODE}" >&2
      cat "${BODY_FILE}" >&2; exit 1 ;;
  esac
  RESULT=$(jq 'if type == "object" and has("data") then .data else . end' "${BODY_FILE}")
  STATUS=$(printf '%s' "${RESULT}" | jq -r '.status // empty')

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

Parameters

Task Submission Parameters

Request Parameters

Every model also accepts the global request options enable_sync_mode and enable_base64_output; see their pages for details.

Response Parameters

Result Request Parameters

Result Response Parameters