图转提示词Image to Prompt
上传参考图，自动分析配色、构图、光线倾向，生成结构化描述填入锚点。Upload a reference image, auto-analyze palette, composition, lighting, generate structured description into anchor.
把 ai-draw-cue-word-project v2.8 的提示词工程资产（角色锚点 · 五大家族语法 · 负面词库 · 渲染前检查 · 视频生成 · 3D 生成 · LoRA 管理 · 多角色空间 · 对话编辑链 · 分镜骨架 · 漫剧模式）变成可交互工具：填一次角色设定，按模型家族自动拼装提示词，渲染前逐项过关。Turn the ai-draw-cue-word-project v2.8 prompt engineering assets (character anchors · five syntax families · negative word bank · pre-render checklist · video generation · 3D generation · LoRA management · multi-character spatial relations · conversational edit chain · storyboard skeleton · comic mode) into interactive tools: fill in a character once, auto-assemble prompts per model family, and pass every check before rendering.
核心设定 → 自动生成角色锚点卡（含权重建议）。锚点是跨模型保持角色一致的根：核心身份锚点 ≥1.6，气质标签 ≥1.3，主描述 0–2。Core setup → auto-generate an anchor card (with weight guidance). Anchors are the root of cross-model consistency: identity anchor ≥1.6, temperament tag ≥1.3, main description 0–2.
保存、复用企业 IP 角色锚点，支持 JSON 导入/导出，团队共享预设。Save and reuse enterprise IP character anchors; JSON import/export; team-shared presets.
自动记录所有提示词生成行为，满足企业溯源合规要求。记录不可手动删除，完整追溯链永久保留。Automatically log all prompt generation behaviors, meet enterprise traceability compliance requirements. Logs cannot be manually deleted, full trace chain is permanently retained.
基于当前锚点，一次性生成多模型提示词，导出 CSV 直接用于下游管线。Generate multi-model prompts from current anchor, export CSV for downstream pipelines.
上传参考图，自动分析配色、构图、光线倾向，生成结构化描述填入锚点。Upload a reference image, auto-analyze palette, composition, lighting, generate structured description into anchor.
基于当前锚点与构图设置，自动生成多版本变体提示词，一键导出。Auto-generate multi-variant prompts from current anchor and composition, one-click export.
面向集成商与提示词工具厂商——证明你的产品遵循 SPL Prompt Specification 标准，可在产品上展示认证徽章。For integrators and prompt tool vendors: prove your product follows the SPL Prompt Specification and display the certification badge.
同一份角色设定 + 构图选择 → 按所选模型家族语法自动拼装。语法差异由引擎处理：自然语言流禁权重括号；MJ 锚点前置 + 尾部参数；SD 加权标签；国产 API 中文段落。Same character setup + composition choices → auto-assembled per the selected model family's syntax. Engine handles differences: NL flows forbid weight brackets; MJ anchors first + trailing params; SD weighted tags; Chinese API paragraphs.
选择模型家族并填写角色设定后，点击「生成提示词」。Select a model family and fill in the character setup, then click Generate Prompt.
9 大失败类别逐项勾选 → 自动汇总负面词清单。每类自带推荐权重与「失败 → 修复」指引。Check any of the 9 failure categories → auto-compile a negative prompt list. Each category carries a recommended weight and a fail→fix guide.
勾选失败类别后生成负面词清单。Check failure categories to generate a negative prompt list.
13 项 P0–P2 验证：P0 必须全过，P1 建议全过，P2 尽力满足。全部打勾后再点生成。13 P0–P2 checks: P0 must all pass, P1 should all pass, P2 best effort. Tick everything before generating.
同一功能在四大模型家族的实现差异 —— 选模型前先看这里，避免用错语法。How the same capability differs across the four model families — check here before choosing a model.
| 功能Feature | MJ / Niji | Stable DiffusionStable Diffusion | 自然语言族 (Flux / DALL·E / GPT-4o / Gemini)Natural Language (Flux / DALL·E / GPT-4o / Gemini) | 国产 API 族 (即梦 / 可灵 / 豆包 / 通义万相)Chinese API (Jimeng / Kling / Doubao / Tongyi) |
|---|
6 大视频模型语法对照 + 通用红线。镜头运动单方向、锚点句全程冻结、禁止场景切换换装。6 video model syntax comparison + universal red lines. Single-direction camera movement, anchor sentence frozen throughout, no scene/outfit switching.
| 模型家族Model Family | 适用场景Use Case | 语法 & 镜头运动Syntax & Camera | 时间结构Time Structure | 时长 / 帧率Duration / FPS | 一致性策略Consistency Strategy | 排除处理Exclusions |
|---|
选模型 → 填主体/场景/镜头 → 自动拼装视频提示词。Select model → fill subject/scene/camera → auto-assemble video prompt.
选择模型、填写内容后点击「生成视频提示词」。Select model, fill content, then click Generate Video Prompt.
6 款 3D 模型输入输出与角色一致性对照。多视角参考锁定 360 度身份，单视角仅锁正面。6 3D model I/O and character consistency comparison. Multi-view references lock 360-degree identity; single-view locks front only.
| 模型Model | 输入格式Input Format | 输出格式Output Format | 纹理Texture | 拓扑质量Topology Quality | 角色一致性Character Consistency | 最佳用途Best Use | 局限Limitations |
|---|
5 种 LoRA 类型 + 堆叠规则 + 与权重系统交互。单 LoRA >1.0 致伪影，总权重 >2.0 致模型失稳。5 LoRA types + stacking rules + weight system interaction. Single LoRA >1.0 causes artifacts, total weight >2.0 destabilizes model.
| LoRA 类型LoRA Type | 用途Purpose | 触发词Trigger Word | 权重范围Weight Range | 冲突检测Conflict Detection | 组合规则Combination Rule | 备注Notes |
|---|
3+ 角色场景的空间排布、深度分层、视线链、遮挡规则、比例透视。三角排布最稳；禁面部遮挡。Spatial layout, depth layering, gaze chains, occlusion rules, proportion perspective for 3+ character scenes. Triangular layout is most stable; no face occlusion.
| 维度Dimension | 说明Description | 示例（3 角色场景）Example (3-char scene) | 负面词Negative | 备注Notes |
|---|
逐轮单点修正：每轮只改一个元素，锚点句全程冻结。3 轮连续修正失败 → 退出对话，从 Round 1 重新生成。Per-round single-point fix: change one element per round, anchor sentence frozen throughout. 3 consecutive failures → exit dialogue, regenerate from Round 1.
| 轮次Round | 动作类型Action | 指令模板Template | 锚点边界Anchor Boundary | 示例Example | 失败处理Failure Handling |
|---|
10 种面板角色 + 3 条镜头连续性规则 + 4 种页面排版。每行含提示词拼装规则、连续性检查项、常见失败。用户只填故事和对话，工具出结构、出提示词、出布局。10 panel roles + 3 camera continuity rules + 4 page layouts. Each row has prompt assembly rules, continuity checks, common failures. Users fill story and dialogue; the tool outputs structure, prompts, and layout.
v2.8 漫剧工作流闭环：面板排布规则 · 对话气泡定位 · 逐帧时序一致性验收。填故事 → 出布局 → 出气泡 → 出验收清单。v2.8 comic workflow: panel layout rules · dialogue bubble positioning · per-frame temporal consistency validation. Fill story → get layout → get bubbles → get checklist.
填写故事大纲并选择排版类型后，点击「生成漫剧布局」。Fill in the story outline and select a layout type, then click Generate Comic Layout.
自定义分镜数量、每镜独立设置镜头/打光/构图，批量导出分镜提示词。支持故事导入、分镜重排、一键生成全系列提示词。Customize number of shots, set camera/lighting/composition independently per shot, batch export storyboard prompts. Support story import, shot reordering, one-click generate full series prompts.