qianwen-image-generation

name	qianwen-image-generation
description	[QianWen] Generate and edit images using Wan and Qwen Image models. Supports text-to-image, image editing (style transfer, subject consistency, text rendering), and interleaved text-image output. TRIGGER when: user wants to create illustrations, product images, artistic designs, posters, text-to-image generation, edit/transform existing images, apply style transfer, generate images based on reference photos, interleaved text-image content, mentions Wan/Qwen Image models/AI art creation, or explicitly invokes this skill by name (e.g. use qianwen-image-generation). DO NOT TRIGGER when: user wants to understand/analyze existing images or OCR (use qianwen-vision), video generation (use qianwen-video-generation), text-only tasks.
compatibility	Requires Python 3.9+ and curl. Cursor: auto-loaded. Claude Code: read this skill's SKILL.md before first use.

Agent setup: If your agent doesn't auto-load skills (e.g. Claude Code), see agent-compatibility.md once per session.

Qwen Image Generation

Generate and edit images using Wan and Qwen Image models. Supports text-to-image, reference-image editing (style transfer, subject consistency, multi-image composition, text rendering), and interleaved text-image output. This skill is part of QianWen-AI/qianwen-ai.

Skill directory

Use this skill's internal files to execute and learn. Load reference files on demand when the default path fails or you need details.

Location	Purpose
scripts/image.py	Default execution — sync/async, upload, download
references/execution-guide.md	Fallback: curl (sync/async), code generation
references/prompt-guide.md	Prompt formulas, style keywords, negative_prompt, prompt_extend decision
references/api-guide.md	API supplement
references/sources.md	Official documentation URLs
references/agent-compatibility.md	Agent self-check: register skills in project config for agents that don't auto-load

Security

NEVER output any API key or credential in plaintext. Always use variable references ($DASHSCOPE_API_KEY in shell, os.environ["DASHSCOPE_API_KEY"] in Python). Any check or detection of credentials must be non-plaintext: report only status (e.g. "set" / "not set", "valid" / "invalid"), never the value. Never display contents of .env or config files that may contain secrets.

When the API key is not configured, NEVER ask the user to provide it directly. Instead, help create a .env file with a placeholder (DASHSCOPE_API_KEY=sk-your-key-here) and instruct the user to replace it with their actual key from the QianWen Console. Only write the actual key value if the user explicitly requests it.

Key Compatibility

Scripts require a standard QianWen API key (sk-...). Token Plan 团队版 keys (sk-sp-...) target a different endpoint (token-plan.cn-beijing.maas.aliyuncs.com) and cannot be used by these scripts. Token Plan does include 4 image generation models (qwen-image-2.0, qwen-image-2.0-pro, wan2.7-image, wan2.7-image-pro), but they are accessed only through interactive AI tools' Skill / Slash Command / Agent mechanism against a dedicated multimodal-generation endpoint — not via this skill's standard DashScope flow. Standard sk- key required for this skill. The script detects sk-sp- keys at startup and prints a warning. If qianwen-ops-auth is installed, see its references/tokenplan.md for full details.

Mode Selection Guide

User Want	Mode	Model
Generate image from text only	t2i	wan2.6-t2i (default), or wan2.7-image / wan2.7-image-pro
Open-source / lowest-cost text-to-image	t2i	z-image-turbo (sync-only; no n; no reference images)
Edit image / apply style transfer based on 1–4 reference images	image-edit	wan2.7-image-pro / wan2.7-image / wan2.6-image
Subject consistency: generate new images maintaining subject from references	image-edit	wan2.7-image-pro / wan2.7-image / wan2.6-image
Multi-image composition: combine style from one image, background from another	image-edit	wan2.7-image-pro / wan2.7-image / wan2.6-image
Single-image editing preserving subject consistency	i2i	wan2.5-i2i-preview
Multi-image fusion: place object from one image into another scene	i2i	wan2.5-i2i-preview
Interleaved text-image output (e.g., tutorials, step-by-step guides)	interleave	wan2.6-image
Fast text-to-image drafts	t2i	wan2.2-t2i-flash
Edit text within images, precise element manipulation	image-edit	qwen-image-2.0-pro
Multi-image fusion with realistic textures	image-edit	qwen-image-2.0-pro
Posters / complex Chinese+English text rendering	t2i	qwen-image-2.0-pro
Text-to-image with fixed aspect ratios (batch)	t2i	qwen-image-plus / qwen-image-max

Model Selection

Wan Series (default)

Model	Use Case
wan2.6-t2i	Recommended for text-to-image — sync + async, best quality
wan2.7-image-pro	Multi-function (4K support) — text-to-image, image editing (0–9 images), sequential multi-image, interactive editing (bbox), thinking mode, color palette. Max 4K for t2i, 2K for editing
wan2.7-image	Multi-function (faster) — same as pro but max 2K, no 4K support
wan2.6-image	Image editing (NOT for pure text-to-image) — requires reference_images or enable_interleave: true. Style transfer, subject consistency (1–4 images), interleaved text-image output, 2K
wan2.5-i2i-preview	Image editing — single-image editing with subject consistency, multi-image fusion (up to 3 images), async-only
wan2.5-t2i-preview	Preview — free size within constraints
wan2.2-t2i-flash	Fast — lower latency
wan2.2-t2i-plus	Professional — improved stability

Other

Model	Use Case
z-image-turbo	Open-source SOTA T2I — sync-only; single text content per message; no n parameter (server returns 400 if provided); no reference images. Parameters: size, prompt_extend, seed.

Qwen Image Series

Model	Use Case
qwen-image-2.0-pro	Fused generation + editing — text rendering, realistic textures, multi-image (1–3 input, 1–6 output)
qwen-image-2.0	Accelerated generation + editing
qwen-image-edit-max	Image editing — 1–6 output images
qwen-image-edit-plus	Image editing — 1–6 output images
qwen-image-edit	Image editing — 1 output image only
qwen-image-plus	Text-to-image — fixed resolutions only (async)
qwen-image-max	Text-to-image — fixed resolutions only

Qwen Image editing models (qwen-image-2.0-pro, qwen-image-2.0, qwen-image-edit-max/plus/edit) use the same sync endpoint as wan2.6-image (/multimodal-generation/generation) with messages format. They support text editing in images, element add/delete/replace, style transfer, and multi-image fusion (1–3 input images). Size range: 512x512 to 2048x2048. qwen-image-2.0-pro and qwen-image-2.0 also support pure text-to-image (no reference images needed).

Qwen Image text-to-image models (qwen-image-plus, qwen-image-max) use a different endpoint (/text2image/image-synthesis) with input.prompt format (async-only). They support only 5 fixed resolutions: 1664*928, 1472*1104, 1328*1328, 1104*1472, 928*1664.

Choosing between wan2.6-image and wan2.5-i2i-preview for image editing:

• wan2.6-image supports up to 4 images, higher resolution (2K), interleaved text-image output, and sync mode. Use for multi-image style composition, interleaved tutorials.
• wan2.5-i2i-preview uses a simpler prompt-only editing interface (no messages format), supports up to 3 images, async-only. Use for straightforward single-image edits and multi-image object fusion.

1. User specified a model → use directly.
2. Consult the qianwen-model-selector skill when model choice depends on requirement, scenario, or pricing.
3. Text-to-image (prompt only, no reference images) → use wan2.6-t2i (default) or wan2.7-image / wan2.7-image-pro (multi-function, higher quality). NEVER use wan2.6-image for pure text-to-image — it will error without reference images or enable_interleave: true.
4. Reference images / image editing / interleaved output → wan2.7-image-pro (recommended), wan2.7-image, or wan2.6-image.

⚠️ Important: The model list above is a point-in-time snapshot and may be outdated. Model availability changes frequently. Always check the official model list for the authoritative, up-to-date catalog before making model decisions.

Model details: For more information about a specific model, direct the user to its detail page: https://www.qianwenai.com/models/<model-name> (replace <model-name> with the exact model ID, e.g. wan2.7-image-pro → https://www.qianwenai.com/models/wan2.7-image-pro). NEVER modify or guess the model name in the URL.

Dynamic model queries: If the qianwen-model-selector skill or QianWen CLI (qianwen models info <model>) is available, use it for real-time model data. CLI requires authentication — see the qianwen-usage skill for login flow.

Execution

⚠️ Multiple artifacts: When generating multiple files in a single session, you MUST append a numeric suffix to each filename (e.g. out_1.png, out_2.png) to prevent overwrites.

Prerequisites

• API Key: Check that DASHSCOPE_API_KEY (or QIANWEN_API_KEY) is set using a non-plaintext check only (e.g. in shell: [ -n "$DASHSCOPE_API_KEY" ]; report only "set" or "not set", never the key value). If not set: run the qianwen-ops-auth skill if available; otherwise guide the user to obtain a key from QianWen Console and set it via .env file (echo 'DASHSCOPE_API_KEY=sk-your-key-here' >> .env in project root or current directory) or environment variable. The script searches for .env in the current working directory and the project root. Skills may be installed independently — do not assume qianwen-ops-auth is present.
• Python 3.9+ (stdlib only, no pip install needed)

Environment Check

Before first execution, verify Python is available:

python3 --version  # must be 3.9+

If python3 is not found, try python --version or py -3 --version. If Python is unavailable or below 3.9, skip to Path 2 (curl) in execution-guide.md.

Default: Run Script

Script path: Scripts are in the scripts/ subdirectory of this skill's directory (the directory containing this SKILL.md). You MUST first locate this skill's installation directory, then ALWAYS use the full absolute path to execute scripts. Do NOT assume scripts are in the current working directory. Do NOT use cd to switch directories before execution.

Execution note: Run all scripts in the foreground — wait for stdout; do not background.

Discovery: Run python3 <this-skill-dir>/scripts/image.py --help first to see all available arguments.

# Text-to-image (wan2.6-t2i, default)
python3 <this-skill-dir>/scripts/image.py \
  --request '{"prompt":"A cozy flower shop with wooden door"}' \
  --output output/qianwen-image-generation/images/out.png \
  --print-response

# Image editing with reference images (wan2.6-image)
python3 <this-skill-dir>/scripts/image.py \
  --model wan2.6-image \
  --request '{"prompt":"Apply watercolor painting style to this photo","reference_images":["https://img.alicdn.com/imgextra/i1/NotRealJustExample/photo.jpg"],"n":1,"size":"1K"}' \
  --output output/qianwen-image-generation/images/out.png \
  --print-response

More examples (interleaved output, wan2.5-i2i, qwen-image-2.0-pro, qwen-image-plus): See execution-guide.md

Argument	Description
--request '{...}'	JSON request body
--file path.json	Load request from file
--async	Force async mode (required for wan2.5 and older; auto-enabled for qwen-image-plus/max and interleaved output)
--model ID	Override model (wan2.6-t2i default; see model list in help)
--output path	Save image to path (or directory for multi-image output)
--print-response	Print response JSON to stdout

Verify Result

• Exit code 0 + output contains valid JSON with output.results or output.task_id → success
• Non-zero exit, HTTP error, empty response, or error JSON → fail
• Async: submission must return output.task_id; poll must reach task_status: SUCCEEDED
• Post-execution check: Verify the output file exists and has non-zero size (ls -la <output_path>)
• MANDATORY — stderr signal check: After confirming the result, scan the command's stderr output for [ACTION_REQUIRED] or [UPDATE_AVAILABLE]. If either signal is present, you MUST follow the instructions in Update Check below before responding to the user.

On Failure

If the script fails, match the error output against the diagnostic table below to determine the resolution. If no match, read execution-guide.md for alternative paths: curl commands (Path 2 — sync and async), code generation (Path 3), and autonomous resolution (Path 5).

If Python is not available at all → skip directly to Path 2 (curl) in execution-guide.md.

Error Pattern	Diagnosis	Resolution
command not found: python3	Python not on PATH	Try python or py -3; install Python 3.9+ if missing
Python 3.9+ required	Script version check failed	Upgrade Python to 3.9+
SyntaxError near type hints	Python < 3.9	Upgrade Python to 3.9+
QIANWEN_API_KEY/DASHSCOPE_API_KEY not found	Missing API key	Obtain key from QianWen Console; add to .env: echo 'DASHSCOPE_API_KEY=sk-...' >> .env; or run qianwen-ops-auth if available
HTTP 401	Invalid or mismatched key	Run qianwen-ops-auth (non-plaintext check only); verify key is valid
SSL: CERTIFICATE_VERIFY_FAILED	SSL cert issue (proxy/corporate)	macOS: run Install Certificates.command; else set SSL_CERT_FILE env var
URLError / ConnectionError	Network unreachable	Check internet; set HTTPS_PROXY if behind proxy
HTTP 429	Rate limited	Wait and retry with backoff
HTTP 5xx	Server error	Retry with backoff
PermissionError	Can't write output	Use --output to specify writable directory

Quick Reference

Request Fields (Common)

Field	Type	Description
prompt	string	Text description of the image to generate (required)
negative_prompt	string	Content to avoid in the image (max 500 chars)
size	string	Resolution — 1280*1280 (t2i default), 1K/2K or width*height (wan2.6-image)
seed	int	Random seed for reproducibility [0, 2147483647]
model	string	wan2.6-t2i (default) or other Wan model
prompt_extend	bool	Enable prompt rewriting (default: true; image editing mode only)

Request Fields (wan2.7-image-pro / wan2.7-image — Multi-function)

Field	Type	Description
reference_images	string[]	0–9 image URLs or local paths
reference_image	string	Single image URL/path (shorthand)
size	string	1K, 2K (default), or 4K (pro only, t2i mode). Or pixel dimensions
enable_sequential	bool	true: sequential multi-image mode (n=1–12). false (default): single/batch mode (n=1–4)
n	int	Images to generate. Sequential mode: 1–12 (default 12). Non-sequential: 1–4 (default 4). Billed per image.
thinking_mode	bool	Enable enhanced reasoning for better quality (default: true). Only for t2i (no images, non-sequential)
bbox_list	List[List[List[int]]]	Interactive editing regions. Format: [[[x1,y1,x2,y2],...], ...]. List length = image count. Empty [] for images without edits
color_palette	array	Custom color theme (3–10 colors). Each: {"hex":"#C2D1E6","ratio":"23.51%"}. Sum of ratios = 100%. Non-sequential mode only
watermark	bool	Add "AI Generated" watermark (default: false)

Note: thinking_mode increases latency but improves quality. enable_sequential generates a coherent image sequence (e.g., same character across scenes).

Request Fields (wan2.6-image — Image Editing)

Field	Type	Description
reference_images	string[]	1–4 image URLs or local paths for editing mode; 0–1 for interleave mode
reference_image	string	Single image URL/path (shorthand; reference_images takes precedence)
enable_interleave	bool	false (default): image editing mode; true: interleaved text-image output
n	int	Number of images to generate in editing mode (1–4, default: 1). Billed per image.
max_images	int	Max images in interleave mode (1–5, default: 5). Billed per image.
watermark	bool	Add "AI Generated" watermark (default: false)

Other Models (wan2.5-i2i, qwen-image-edit, qwen-image-plus/max)

These models have specific parameter requirements:

Model	Key Differences
wan2.5-i2i-preview	async-only, 1–3 images, prompt+images[] format (not messages)
qwen-image-edit-*	1–3 images, n=1–6 (except qwen-image-edit: n=1 only), no interleave
qwen-image-plus/max	async-only, n fixed at 1, 5 fixed resolutions only

Full parameter tables: See api-guide.md for detailed parameters.

Size Reference (wan2.6-image)

• Editing mode: 1K (default, ~1280×1280) or 2K (~2048×2048)
• Interleave mode: pixel dimensions with total pixels in [768×768, 1280×1280]

Common aspect ratios: 1280*1280 (1:1), 960*1280 (3:4), 1280*960 (4:3), 720*1280 (9:16), 1280*720 (16:9)

Response Fields

Field	Description
image_url	URL of generated image (24h validity). Use this when chaining to another skill.
image_urls	Array of all image URLs (multi-image output, wan2.6-image, qwen-image-edit)
image_count	Number of generated images
local_path	Local file path of the downloaded image. Use this for user preview or non-API operations.
local_paths	Array of local file paths (multi-image output)
interleaved_content	Array of {type, text/image} objects (interleave mode)
width / height	Image dimensions
seed	Seed used

API Details

• Sync endpoint (wan2.6-t2i, wan2.6-image editing, qwen-image-edit series): POST /api/v1/services/aigc/multimodal-generation/generation
• Async endpoint (wan2.6 and older t2i): POST /api/v1/services/aigc/image-generation/generation with X-DashScope-Async: enable
• Async endpoint (wan2.5-i2i-preview): POST /api/v1/services/aigc/image2image/image-synthesis with X-DashScope-Async: enable
• Async endpoint (qwen-image-plus, qwen-image-max): POST /api/v1/services/aigc/text2image/image-synthesis with X-DashScope-Async: enable
• wan2.6-t2i resolution: Total pixels in [1280x1280, 1440x1440], aspect ratio [1:4, 4:1]
• wan2.6-image resolution: Editing mode [768x768, 2048x2048]; Interleave mode [768x768, 1280x1280]; aspect ratio [1:4, 4:1]
• Input images (wan2.6-image): JPEG/JPG/PNG/BMP/WEBP, 240–8000px per dimension, ≤10MB
• Local files: Script auto-uploads to DashScope temp storage (oss:// URL, 48h TTL). Pass local paths directly — no manual upload step needed.
• Production: Default temp storage has 48h TTL and 100 QPS upload limit — not suitable for production, high-concurrency, or load-testing. To use your own OSS bucket, set QWEN_TMP_OSS_BUCKET and QWEN_TMP_OSS_REGION in .env, install pip install oss2, and provide credentials via QWEN_TMP_OSS_AK_ID / QWEN_TMP_OSS_AK_SECRET or the standard OSS_ACCESS_KEY_ID / OSS_ACCESS_KEY_SECRET. Use a RAM user with least-privilege (oss:PutObject + oss:GetObject on target bucket only). If qianwen-ops-auth is installed, see its references/custom-oss.md for the full setup guide.
• Interleaved sync: Requires streaming (X-DashScope-Sse: enable + stream: true); use async mode via this script instead

Cross-Skill Chaining

When using generated images as input for another skill (e.g., video-gen i2v, vision analyze):

• Pass image_url directly — do NOT download and re-pass as local path
• All downstream scripts detect URL prefixes (https://, oss://) and pass them through without re-upload
• Use local_path only for user preview or non-API operations (e.g., opening in editor)

Scenario	Use
Feed to another skill (video-gen, vision, image-edit)	image_url (URL)
Show to user / open in editor	local_path (local file)

Error Handling

HTTP	Meaning	Action
401	Invalid or missing API key	Run qianwen-ops-auth if available; else prompt user to set key (non-plaintext check only)
400	Bad request (invalid prompt, size)	Verify parameters and constraints
400 product not activated	Model/product not enabled for this account	The model has not been activated. Direct the user to enable the model then retry
429	Rate limited	Retry with exponential backoff
5xx	Server error	Retry with exponential backoff

Usage & billing: Use the qianwen-usage skill to check usage, free tier quota, and billing directly. Alternatively, the user can visit the QianWen console: Usage Analytics | Pay-as-you-go Billing | Token Plan 团队版 Subscription

NEVER fabricate, guess, or construct usage/billing/console URLs. Only provide the exact links listed in this skill. If a URL is not listed here, do not invent one.

Output Location

Prefer the current working directory. Default subdirectory: ./output/qianwen-image-generation/.

Write prohibition: NEVER write output files into this skill's installation directory or any skills/ hierarchy. All generated content must go to output/ under the current working directory or a user-specified path.

Update Check (MANDATORY Post-Execution)

After script execution, check stderr for action signals. The script automatically detects update status and emits structured signals. You MUST act on them. This section also applies when the user explicitly asks to check for updates.

Signal: [ACTION_REQUIRED] qianwen-update-check skill is not installed

When stderr contains this signal, you MUST present the following options to the user and wait for their response before proceeding. Do NOT silently skip this.

The update-check skill is not installed. It keeps you informed about new QianWen-AI/qianwen-ai features and bug fixes.

1. Install now — I'll run the install command for you
2. Skip for now — do nothing this time
3. Never remind me — permanently suppress this prompt

After the user responds:

• User chooses 1 → run: npx skills add QianWen-AI/qianwen-ai --skill qianwen-update-check -y, then run: python3 <this-skill-scripts-dir>/gossamer.py --dismiss
• User chooses 2 → run: python3 <this-skill-scripts-dir>/gossamer.py --dismiss, then continue normally
• User chooses 3 → run: python3 <this-skill-scripts-dir>/gossamer.py --never-install, then continue normally

Signal: [UPDATE_AVAILABLE]

When stderr contains this signal, you MUST append a brief update notice to your response, including the version info and the update command shown in the stderr output.

No signal in stderr

If stderr contains neither [ACTION_REQUIRED] nor [UPDATE_AVAILABLE], no action is needed — the skill is installed and up to date (or cached within 24h).

Explicit user request

When the user explicitly asks to check for updates (e.g. "check for updates", "check version"):

1. Look for qianwen-update-check/SKILL.md in sibling skill directories.
2. If found — run: python3 <qianwen-update-check-dir>/scripts/check_update.py --print-response and report the result.
3. If not found — present the install options above.

References

• execution-guide.md — Fallback paths (curl sync/async, code generation, autonomous)
• api-guide.md — API supplementary guide
• sources.md — Official documentation URLs