Meshy 3D — Generation + Printing

Directly communicate with the Meshy AI API to generate and print 3D assets. Covers the complete lifecycle: API key setup, task creation, exponential backoff polling, downloading, multi-step pipelines, and 3D print preparation with slicer integration.

---

SECURITY MANIFEST

Environment variables accessed:

• - MESHY_API_KEY — API authentication token sent in HTTP Authorization: Bearer header only. Never logged, never written to any file except .env in the current working directory when explicitly requested by the user.

External network endpoints:

• - https://api.meshy.ai — Meshy AI API (task creation, status polling, model/image downloads)

File system access:

• - Read: .env in the current working directory only (API key lookup)
• Write: .env in the current working directory only (API key storage, only on user request)
• Write: ./meshy_output/ in the current working directory (downloaded model files, metadata)
• Read: files explicitly provided by the user (e.g., local images passed for image-to-3D conversion), accessed only at the exact path the user specifies
• No access to home directories, shell profiles, or any path outside the above

Data leaving this machine:

• - API requests to api.meshy.ai include the MESHY_API_KEY in the Authorization header and user-provided text prompts or image URLs. No other local data is transmitted. Downloaded model files are saved locally only.

---

IMPORTANT: First-Use Session Notice

When this skill is first activated in a session, inform the user:

All generated files will be saved to meshy_output/ in the current working directory. Each project gets its own folder ({YYYYMMDD_HHmmss}_{prompt}_{id}/) with model files, textures, thumbnails, and metadata. History is tracked in meshy_output/history.json.

This only needs to be said once per session.

---

IMPORTANT: File Organization

All downloaded files MUST go into a structured meshy_output/ directory in the current working directory. Do NOT scatter files randomly.

• - Each project: INLINECODE13
• Chained tasks (preview → refine → rig) reuse the same INLINECODE14
• Track tasks in metadata.json per project, and global INLINECODE16
• Auto-download thumbnails alongside models

---

IMPORTANT: Shell Command Rules

Use only standard POSIX tools. Do NOT use rg, fd, bat, exa/eza.

---

IMPORTANT: Run Long Tasks Properly

Meshy generation takes 1–5 minutes. Write the entire create → poll → download flow as ONE Python script and execute in a single Bash call. Use python3 -u script.py for unbuffered output. Tasks sitting at 99% for 30–120s is normal finalization — do NOT interrupt.

---

Step 0: API Key Detection (ALWAYS RUN FIRST)

Only check the current session environment and the .env file in the current working directory. Do NOT scan home directories or shell profile files.

CODEBLOCK0

Decision After Detection

• - Key found → Proceed to Step 1.
• Key NOT found → Go to Step 0a.
• Python requests missing → Run pip install requests.

---

Step 0a: API Key Setup (Only If No Key Found)

Tell the user:

To use the Meshy API, you need an API key:

1. 1. Go to https://www.meshy.ai/settings/api
2. Click "Create API Key", name it, and copy the key (starts with msy_)
3. The key is shown only once — save it somewhere safe

Note: API access requires a Pro plan or above. Free-tier accounts cannot create API keys.

Once the user provides the key, set it for the current session and optionally persist to .env:

CODEBLOCK1

To persist the key (current project only):

CODEBLOCK2

Security reminder: The key is stored only in .env in your current project directory. Never commit this file to version control. .env has been automatically added to .gitignore.

---

Step 1: Confirm Plan With User Before Spending Credits

CRITICAL: Before creating any task, present the user with a cost summary and wait for confirmation:

CODEBLOCK3

For multi-step pipelines (text-to-3d → rig → animate), show the FULL pipeline cost upfront.

Note: Rigging automatically includes walking + running animations at no extra cost. Only add Animate (3 credits) for custom animations beyond those.

Intent → API Mapping

User wants to...	API	Endpoint	Credits
3D model from text	Text to 3D	INLINECODE31	20 + 10
3D model from one image

Image to 3D | POST /openapi/v1/image-to-3d | 20–30 | | 3D model from multiple images | Multi-Image to 3D | POST /openapi/v1/multi-image-to-3d | 20–30 | | New textures on existing model | Retexture | POST /openapi/v1/retexture | 10 | | Change mesh format/topology | Remesh | POST /openapi/v1/remesh | 5 | | Add skeleton to character | Auto-Rigging | POST /openapi/v1/rigging | 5 | | Animate a rigged character | Animation | POST /openapi/v1/animations | 3 | | 2D image from text | Text to Image | POST /openapi/v1/text-to-image | 3–9 | | Transform a 2D image | Image to Image | POST /openapi/v1/image-to-image | 3–9 | | Check credit balance | Balance | GET /openapi/v1/balance | 0 | | 3D print a model | → See Print Pipeline section | — | 20 |

---

Step 2: Execute the Workflow

Reusable Script Template

Use this as the base for ALL workflows. It loads the API key securely from environment or .env in the current directory only:

CODEBLOCK4

---

Text to 3D (Preview + Refine)

Append to the template above:

CODEBLOCK5

Note: Only previews from meshy-5 or latest support refine. meshy-6 previews do NOT (API returns 400).

---

Image to 3D

CODEBLOCK6

---

Multi-Image to 3D

CODEBLOCK7

---

Retexture

CODEBLOCK8

---

Remesh / Format Conversion

CODEBLOCK9

---

Auto-Rigging + Animation

When the user asks to rig or animate, the generation step MUST use pose_mode: "t-pose".

CODEBLOCK10

---

Text to Image / Image to Image

CODEBLOCK11

---

3D Printing Workflow

Trigger when the user mentions: print, 3d print, slicer, slice, bambu, orca, prusa, cura, figurine, miniature, statue, physical model, desk toy, phone stand.

Print Pipelines

Text-to-3D Print:

Step	Action	Credits
1	Text to 3D (mode: "preview", no texture)	20
2

Printability check (see checklist) | 0 |
| 3 | Download OBJ | 0 |
| 4 | Open in slicer (direct launch or manual import) | 0 |
| 5 (optional) | Retexture for multi-color | 10 |

Image-to-3D Print:

Step	Action	Credits
1	Image to 3D with INLINECODE47	20
2

Printability check | 0 |
| 3 | Download OBJ | 0 |
| 4 | Open in slicer (direct launch or manual import) | 0 |

Print Download + Slicer Script

Append to the template after task SUCCEEDED:

CODEBLOCK12

INLINECODE48: Default 75mm. Adjust based on user request (e.g. "print at 15cm" → 150.0).

Opening OBJ in slicer: When the user specifies a slicer (e.g. Bambu Studio, OrcaSlicer, Creality Print, PrusaSlicer, Cura), open the downloaded OBJ file directly:

• - macOS: subprocess.run(["open", "-a", "<AppName>", obj_path]) — the OS resolves the app location automatically.
• Windows / Linux: Use shutil.which("<binary_name>") to find the executable in PATH, then subprocess.Popen([exe, obj_path]). If not found, print the file path and instruct manual open.
• No slicer specified: Print the OBJ file path and instruct: File → Import / Open → select .obj file.

Printability Checklist (Manual Review)

Automated printability analysis API is coming soon.

Check	Recommendation
Wall thickness	Min 1.2mm FDM, 0.8mm resin
Overhangs

Keep below 45° or add supports | | Manifold mesh | Watertight, no holes | | Minimum detail | 0.4mm FDM, 0.05mm resin | | Base stability | Flat base or add brim/raft in slicer | | Floating parts | All parts connected or printed separately |

Multi-Color Printing (Manual Guidance)

Automated multi-color API is coming soon.

1. 1. Use Retexture (10 credits) to apply distinct color regions
2. Download OBJ
3. In slicer's color painting tool, assign filament colors to regions
4. Slice with multi-color setup (Bambu AMS, Prusa MMU)

---

Step 3: Report Results

After task succeeds:

1. 1. Downloaded file paths and sizes
2. Task IDs (for follow-up: refine, rig, retexture)
3. Available formats (list model_urls keys)
4. Credits consumed + current balance
5. Suggested next steps:

- Preview done → "Want to refine (add textures)?"
- Model done → "Want to rig this character?"
- Rigged → "Want to apply a custom animation?"
- Any model → "Want to 3D print this?"

---

Error Recovery

HTTP Status	Meaning	Action
401	Invalid API key	Re-run Step 0; ask user to check key
402

Insufficient credits | Show balance, link https://www.meshy.ai/pricing |
| 422 | Cannot process | Explain (e.g., non-humanoid for rigging) |
| 429 | Rate limited | Auto-retry after 5s (max 3 times) |
| 5xx | Server error | Auto-retry after 10s (once) |

Task FAILED messages:

• - "The server is busy..." → retry with backoff (5s, 10s, 20s)
• INLINECODE56 → simplify prompt, retry once

---

Known Behaviors & Constraints

• - 99% stall: Normal finalization (30–120s). Do NOT interrupt.
• Asset retention: Files deleted after 3 days (non-Enterprise). Download immediately.
• PBR maps: Must set enable_pbr: true explicitly.
• Refine: Only meshy-5 / latest previews support refine; meshy-6 does not.
• Rigging: Humanoid bipedal only, polycount ≤ 300,000.
• OBJ for printing: Always download OBJ for slicer compatibility (3MF not yet available from API). If user specifies a slicer, try to open OBJ directly; otherwise print file path for manual import.
• Timestamps: All API timestamps are Unix epoch milliseconds.

---

Execution Checklist

• - [ ] Ran API key detection (Step 0) — checked env var and .env only
• [ ] API key verified (never printed in full)
• [ ] Presented cost summary and got user confirmation
• [ ] Wrote complete workflow as single Python script
• [ ] Ran with python3 -u for unbuffered output
• [ ] Reported file paths, formats, task IDs, and balance
• [ ] Suggested next steps

---

Additional Resources

For the complete API endpoint reference including all parameters, response schemas, and error codes, read reference.md.