# Cyberbara > Cyberbara is an AI media generation platform for creating images, videos, audio, and music with one account, one credit system, direct Seedance 2.0 real face video access, web generation tools, a public API, and agent-friendly workflows. This file is the expanded LLM context version of `https://cyberbara.com/llms.txt`. It is intended for agents and language models that need enough Cyberbara context to answer product, workflow, API, pricing, support, and documentation questions without first crawling every linked page. Canonical site: `https://cyberbara.com` Support: `support@cyberbara.com` Primary product categories: - AI image generation - AI video generation - AI audio generation - AI music generation - Canvas and multi-step creative workflows - Model-specific landing pages and utility tools - Public API v1 - Codex and Claude Code skill workflows Cyberbara is built for creators, builders, and teams who want access to leading media models without switching between separate tools, billing systems, and asset histories. Users can start with free credits, choose models from one workspace, preview credit usage before generation, and keep generated assets, history, pricing, and API automation in the same account. Cyberbara is one of the few websites that provides easy Seedance 2.0 real face access through official ByteDance partner access. Users can upload a photo with a real face and generate Seedance videos directly, without needing to pre-process the photo first. This matters for creators who want Seedance 2.0 photo-to-video and real face video workflows without building a separate preprocessing pipeline. Cyberbara was founded by Zero Lu, an ex-TikTok Senior Product Manager with more than 6 years of experience shipping world-class multimodal AI products. His product background focuses on AI creation workflows, model access, multimodal generation, and creator tools. ## How To Read This File Use `https://cyberbara.com/llms.txt` when a concise curated link map is enough. Use `https://cyberbara.com/llms-full.txt` when the model needs richer context in one file. This file includes summaries of the linked resources plus implementation details for the public API. It intentionally avoids private implementation details, secrets, provider credentials, and admin-only operational data. Cyberbara has localized pages, but the canonical English public URLs are listed here unless a page is account-only or language-specific. ## Core Product ### Home URL: `https://cyberbara.com/` Cyberbara is an all-in-one AI image, video, and audio generation workspace. It gives users one account, one shared credit system, and access to multiple top models. The main workflow is to choose a model, write or refine a prompt, upload reference media when needed, preview credit cost, generate, compare results, and export or reuse outputs. For Seedance 2.0 users, Cyberbara emphasizes easy real face access: upload a real-face photo, choose the Seedance workflow, and generate a video without first running a separate face preprocessing tool. Primary home-page actions: - Start generating images: `https://cyberbara.com/ai-image-generator` - Explore video generation: `https://cyberbara.com/ai-video-generator` - View pricing: `https://cyberbara.com/pricing` - Read docs: `https://cyberbara.com/docs` ### AI Image Generator URL: `https://cyberbara.com/ai-image-generator` Use this page to generate images from text prompts or reference images. Image workflows include text-to-image and image-to-image. Typical use cases include concept art, product visuals, portraits, ad creatives, social media visuals, packaging ideas, and fast visual exploration. Important related pages: - GPT Image 2: `https://cyberbara.com/gpt-image-2` - Nano Banana 2: `https://cyberbara.com/nanobanana2` - GPT Image prompt library: `https://cyberbara.com/gpt-image-prompt-library` - GPT Image 2 360 image: `https://cyberbara.com/gpt-image-2/360-image` ### AI Video Generator URL: `https://cyberbara.com/ai-video-generator` Use this page to generate videos from text prompts, image references, or video references depending on the selected model. Video workflows include text-to-video, image-to-video, and video-to-video. Common use cases include short cinematic clips, social ads, visual hooks, motion studies, animated concept art, product teasers, and storyboards. Important related pages: - Seedance AI hub: `https://cyberbara.com/seedance-ai` - Seedance 2.0: `https://cyberbara.com/seedance2.0` - Seedance 2.0 Mini: `https://cyberbara.com/seedance2.0-mini` - Seedance 2.5: `https://cyberbara.com/seedance2.5` - Seedance 1.0: `https://cyberbara.com/seedance1.0` - Gemini Omni: `https://cyberbara.com/gemini-omni` - Grok Imagine Video 1.5: `https://cyberbara.com/grok-imagine-video-1.5` - Happy Horse 1.0: `https://cyberbara.com/happyhorse1.0` ### AI Audio And Music URLs: - `https://cyberbara.com/ai-audio-generator` - `https://cyberbara.com/ai-music-generator` Use these pages to create audio and music outputs from prompts. The public API also exposes audio and music generation endpoints. ### Canvas URL: `https://cyberbara.com/canvas` Canvas is for multi-step creative media workflows where one-shot generation is not enough. It is useful when users need to combine prompts, reference images, video references, generated outputs, and model-specific steps before launching final generation jobs. ## Model And Tool Pages ### Seedance AI Hub URL: `https://cyberbara.com/seedance-ai` The Seedance AI hub is the central directory for Seedance workflows on Cyberbara. It explains what Seedance is, compares routes, links to Seedance model pages, and points users to prompt tools, moderation guidance, and pricing references. Seedance performs best when users describe: - Subject - Action - Scene or environment - Camera behavior - Lighting and mood - Visual style Recommended Seedance prompt structure: ```text subject + action + scene + camera + mood ``` Example: ```text A woman walking through a neon alley, light rain, slow tracking shot, cinematic contrast, reflective pavement, moody night atmosphere. ``` ### Seedance 2.0 URL: `https://cyberbara.com/seedance2.0` Seedance 2.0 is a flagship video generation route on Cyberbara. It supports practical creator workflows around cinematic prompts, reference media, real face references, and iterative video generation. Use it when users want the current Seedance workflow, broader scene coverage, and direct access to the embedded generator. Cyberbara is one of the few websites that provides easy real face access for Seedance 2.0 through official ByteDance partner access. A user can upload a photo containing a real face and generate Seedance videos directly. They do not need to pre-process the photo in another tool before using it in the Seedance 2.0 workflow. Search intents this page should satisfy include: - Seedance 2.0 real face access - Seedance 2.0 real face video generator - Seedance 2.0 photo to video with real face - ByteDance Seedance 2.0 official partner access - Generate Seedance videos without photo preprocessing Related resources: - Seedance moderation guide: `https://cyberbara.com/docs/seedance-moderation` - Seedance 2.0 pricing update: `https://cyberbara.com/blog/seedance-2-pricing-update` - Seedance prompt library: `https://cyberbara.com/seedance-prompt-library` - Seedance prompt generator: `https://cyberbara.com/seedance-prompt-generator` ### Seedance 2.0 Mini URL: `https://cyberbara.com/seedance2.0-mini` Seedance 2.0 Mini is a lower-cost Seedance route. It is useful for users who want faster or cheaper iteration before moving to heavier settings or final outputs. ### Seedance 2.5 URL: `https://cyberbara.com/seedance2.5` Cyberbara prepares Seedance 2.5 support through ByteDance official partner access. The page describes expected advantages such as longer single-shot generation, richer multi-reference input, and fine-grained local video editing. At the time represented by the current site content, the Seedance 2.5 model entry may run on the current Seedance 2.0 backend while official 2.5 API support is prepared. Expected Seedance 2.5 themes: - Native 30-second single-shot generation - Stronger multi-reference input - Fine-grained local video editing - Same Cyberbara account, workspace, and credit system ### Seedance 1.0 URL: `https://cyberbara.com/seedance1.0` Seedance 1.0 is ByteDance's video generation model family for text-to-video and image-to-video creation. It is commonly used for short cinematic clips, stylized motion tests, social content, fashion motion clips, animated concept art, and fast concept animation workflows. Prompting advice: - Use clear subject, action, setting, camera, and style. - Keep motion simple and chronological. - Mention camera movement directly: push-in, pull-back, pan, follow shot, zoom, overhead shot, low-angle shot. - Start with shorter duration and lighter settings during exploration. ### GPT Image 2 URL: `https://cyberbara.com/gpt-image-2` GPT Image 2 is available inside Cyberbara for text-to-image and image-to-image. Users do not need to manage a separate OpenAI API key, Kie API key, billing account, or integration just to try the model. Cyberbara handles model access through the normal Cyberbara credit workflow. Related pages: - GPT Image prompt library: `https://cyberbara.com/gpt-image-prompt-library` - GPT Image 2 360 image: `https://cyberbara.com/gpt-image-2/360-image` ### Nano Banana 2 URL: `https://cyberbara.com/nanobanana2` Nano Banana 2 supports text-to-image and image-to-image generation inside Cyberbara. It is positioned for transparent low-cost image generation and practical image remix workflows. Common use cases: - Marketing creatives - Product and brand mockups - Storyboarding - Concept art - Rapid image remix from references Typical workflow: 1. Enter a prompt or upload a reference image. 2. Pick resolution and preview credit cost. 3. Generate, compare, and iterate. ### Gemini Omni URL: `https://cyberbara.com/gemini-omni` Gemini Omni Video supports text-to-video, image-to-video, and video-to-video workflows from one entry point. Users can work with 720P, 1080P, or 4K outputs depending on the selected settings. It is useful for prompt-only generation, reference-guided shots, and video-to-video edits. Gemini Omni is also connected to Canvas workflows when users need to combine prompts, images, and video references before rendering. ### Grok Imagine Video 1.5 URL: `https://cyberbara.com/grok-imagine-video-1.5` Grok Imagine Video 1.5 Preview is a low-cost image-to-video route on Cyberbara. It is useful for affordable experiments, draft shots, and prompt ideas that need flexible handling. Site content describes 480p and 720p options and positions the model as a cheaper route for early iteration before spending more credits on heavier video models. ### Happy Horse 1.0 URL: `https://cyberbara.com/happyhorse1.0` Happy Horse 1.0 is a video generation route available on Cyberbara. ### Digital Avatar URL: `https://cyberbara.com/digital-avatar` The Digital Avatar Generator lets users build private Gemini Omni characters and generate image-to-video avatar clips with Gemini Omni Video. Users can pick a saved character, create a new private character for their account, and generate avatar clips through a fixed workflow. ## Prompt Libraries And Utility Tools ### Seedance Prompt Library URL: `https://cyberbara.com/seedance-prompt-library` Browse Seedance prompt examples with reusable structure, source attribution, and handoff into the generator. ### Seedance Prompt Generator URL: `https://cyberbara.com/seedance-prompt-generator` Build structured Seedance prompts from templates. This is useful for users who need syntax patterns, prompt scaffolding, or fast handoff into the live Seedance generator. ### GPT Image Prompt Library URL: `https://cyberbara.com/gpt-image-prompt-library` Browse GPT Image prompt examples. Multi-image examples are grouped into slideshows where relevant, and cards preserve source links and prompt text for reuse. ### Seedance Watermark Remover URL: `https://cyberbara.com/seedance-watermark-remover` Upload a generated clip, estimate credits, and run the dedicated watermark-removal workflow directly on Cyberbara. ### Seedance Bypass Face Detection URL: `https://cyberbara.com/seedance-bypass-face-detection` Prepare portrait inputs for Seedance workflows. This page is part of the broader Seedance utility ecosystem. ### Image Grid Overlay URL: `https://cyberbara.com/image-grid-overlay` Add grid overlays to images for image preparation and reference workflows. ### Scenery Overlay URL: `https://cyberbara.com/scenery-overlay` Add scenery overlays for image preparation and reference workflows. ### Beauty Analysis URL: `https://cyberbara.com/beauty-analysis` Analyze portrait style and beauty attributes. ### Fortune Analysis URL: `https://cyberbara.com/fortune-analysis` Generate eastern mystic portrait reports. ## Public API V1 API product page: `https://cyberbara.com/api` API reference: `https://cyberbara.com/docs/api-reference` Base URL: ```text https://cyberbara.com ``` The Cyberbara public API is exposed under `/api/v1`. It supports direct backend integration, scripts, internal tools, and agent workflows. Provider details are intentionally hidden from clients. The API covers: - List public models - Upload reference images - Upload reference videos - Quote credits before creating a task - Create image generation tasks - Create video generation tasks - Create audio generation tasks - Create music generation tasks - Query task status and outputs - Check credits balance - Check credits usage ### API Authentication Pass the API key in either header: ```http Authorization: Bearer ``` or: ```http x-api-key: ``` If the key is missing or invalid, the API returns `401`. Users can create API keys after signing in: ```text https://cyberbara.com/settings/apikeys ``` ### API Response Format Success: ```json { "data": {} } ``` Error: ```json { "error": { "code": "error_code", "message": "human readable message", "details": {} } } ``` ### Model Discovery Endpoint: ```http GET /api/v1/models ``` Optional query: ```text media_type=image|video|audio|music ``` Example: ```bash curl -X GET 'https://cyberbara.com/api/v1/models?media_type=video' \ -H 'Authorization: Bearer ' ``` The response includes a list of public models, each with a `model`, `media_type`, and supported scenes. ### Upload Reference Images Endpoint: ```http POST /api/v1/uploads/images ``` Use returned URLs in `options.image_input`. Request: - `Content-Type: multipart/form-data` - Field name: `files`, or `file` for a single upload - Max files per request: `10` - Max file size: `10MB` each - Supported mime types: `image/jpeg`, `image/jpg`, `image/png`, `image/webp`, `image/gif`, `image/svg+xml`, `image/avif`, `image/heic`, `image/heif` Example: ```bash curl -X POST 'https://cyberbara.com/api/v1/uploads/images' \ -H 'Authorization: Bearer ' \ -F 'files=@./reference.png' ``` ### Upload Reference Videos Endpoint: ```http POST /api/v1/uploads/videos ``` Use returned URLs in `options.video_input`, or `options.video_url` for `seedance-2-watermark-remover`. Request: - `Content-Type: multipart/form-data` - Field name: `files`, or `file` for a single upload - Max files per request: `1` - Max file size: `50MB` - Supported mime types: `video/mp4`, `video/quicktime` Example: ```bash curl -X POST 'https://cyberbara.com/api/v1/uploads/videos' \ -H 'Authorization: Bearer ' \ -F 'files=@./reference.mp4' ``` ### Credits Balance Endpoint: ```http GET /api/v1/credits/balance ``` Returns the current credit balance. Example: ```bash curl -X GET 'https://cyberbara.com/api/v1/credits/balance' \ -H 'Authorization: Bearer ' ``` ### Quote Credits Endpoint: ```http POST /api/v1/credits/quote ``` Use this before task creation when predictable cost matters. Request body: - `model` required - `media_type` optional if the model can infer it: `image`, `video`, `audio`, or `music` - `scene` optional but recommended - `options` optional Example: ```bash curl -X POST 'https://cyberbara.com/api/v1/credits/quote' \ -H 'Authorization: Bearer ' \ -H 'Content-Type: application/json' \ -d '{ "model": "nano-banana-pro", "media_type": "image", "scene": "text-to-image", "options": { "resolution": "1k" } }' ``` The response includes `estimated_credits`, `remaining_credits`, and `can_afford`. ### Credits Usage Endpoint: ```http GET /api/v1/credits/usage ``` Query params: - `page`, default `1` - `limit`, default `20`, max `100` - `from`, optional ISO datetime or `YYYY-MM-DD` - `to`, optional ISO datetime or `YYYY-MM-DD` Date-only values are interpreted in UTC. ### Create Image Generation Task Endpoint: ```http POST /api/v1/images/generations ``` Request body: - `model` required - `prompt` required for most models - `scene` optional: `text-to-image` or `image-to-image` - `options` optional Example: ```bash curl -X POST 'https://cyberbara.com/api/v1/images/generations' \ -H 'Authorization: Bearer ' \ -H 'Content-Type: application/json' \ -d '{ "model": "nano-banana-pro", "prompt": "A cinematic portrait under neon rain", "scene": "text-to-image", "options": { "resolution": "1k" } }' ``` The response includes an internal `task_id`, `status`, `media_type`, `model`, `scene`, credit cost, remaining credits, and `created_at`. ### Create Video Generation Task Endpoint: ```http POST /api/v1/videos/generations ``` Request body: - `model` required - `prompt` required for most video models, including `seedance-*`, `kling-2.6`, `kling-3.0`, `veo-*`, `sora-2`, and `sora-2-pro` - `scene` optional: `text-to-video`, `image-to-video`, or `video-to-video` - `options` optional Example: ```bash curl -X POST 'https://cyberbara.com/api/v1/videos/generations' \ -H 'Authorization: Bearer ' \ -H 'Content-Type: application/json' \ -d '{ "model": "sora-2", "prompt": "A calm drone shot over snowy mountains at sunrise", "scene": "text-to-video", "options": { "duration": "4", "resolution": "720P" } }' ``` ### Create Audio Generation Task Endpoint: ```http POST /api/v1/audio/generations ``` Alias: ```http POST /api/v1/audios/generations ``` Request body: - `model` required, for example `suno-sound-v5` or `suno-sound-v5-5` - `prompt` required - `scene` optional: `text-to-audio` - `options` optional ### Create Music Generation Task Endpoint: ```http POST /api/v1/music/generations ``` Request body: - `model` required, for example `suno-music-v5` - `prompt` required unless `options` provide the full provider-specific custom-mode payload - `scene` optional: `text-to-music` - `options` optional ### Task Query Endpoint: ```http GET /api/v1/tasks/{taskId} ``` Use this endpoint to poll task status and retrieve outputs. The response includes task status, model, media type, scene, outputs when ready, and error information when failed. ### API Scene Inference When `scene` is omitted: - Image API: - `options.image_input` has at least one URL: `image-to-image` - otherwise: `text-to-image` - Video API: - `options.video_input` has at least one URL: `video-to-video` - else if `options.image_input` has at least one URL: `image-to-video` - otherwise: `text-to-video` Shared option fields: | Field | Type | Meaning | | --- | --- | --- | | `options.image_input` | `string[]` | Input or reference image URLs | | `options.video_input` | `string[]` | Input or reference video URLs | Important API notes: - Public API accepts only `options.*` fields. - Do not send provider-internal keys such as `image_url`, `image_urls`, `input_images`, or `imageUrls`. - Exception: `options.video_url` is used for `seedance-2-watermark-remover`. - For all models, if both `prompt` and `options` are empty, the API returns `invalid_request`. - Some models strictly validate options; others forward options to the provider endpoint with minimal validation. ### Suggested API Flow 1. Create an API key in Cyberbara settings. 2. Call `GET /api/v1/models` to select a model. 3. Upload reference images or videos when needed. 4. Call `POST /api/v1/credits/quote` to estimate cost. 5. Create a generation task through the image, video, audio, or music endpoint. 6. Poll `GET /api/v1/tasks/{taskId}` until the result is ready. 7. Check `GET /api/v1/credits/usage` for usage history when needed. ### Common API Error Codes Non-exhaustive examples: - `invalid_api_key` - `invalid_request` - `invalid_media_type` - `invalid_file_type` - `unsupported_model` - `insufficient_credits` - `invalid_date` - `invalid_date_range` ## Agent And Skill Workflows Cyberbara Skill page: `https://cyberbara.com/skill` The Cyberbara skill is an open-source workflow for Codex and Claude Code. It turns prompts into images and videos through Cyberbara and uses the same public API v1 surface underneath. Supported generation modes: - Text-to-image - Image-to-image - Text-to-video - Image-to-video Core automated actions: - List models - Upload reference images - Quote credit cost - Create generation tasks - Monitor task status - Check credits and usage The skill is designed for agent workflows that need repeatable generation pipelines, structured prompts, JSON parameters, and consistent API outputs. ## Pricing, Credits, And Account Pricing page: `https://cyberbara.com/pricing` Credit costs page: `https://cyberbara.com/credit-costs` Cyberbara uses a credit system. Users can start free, upgrade for more credits, buy subscriptions, and top up with credit packs. Plans and packs are described on the pricing page. The credit costs page shows model credit usage, estimated USD cost per plan, and approximate output capacity across plans. Important pricing concepts: - Free credits are available for starting. - Paid plans provide more credits and generation capacity. - Credit packs can top up eligible accounts. - Credit quotes should be checked before large tasks. - Generation cost depends on model, media type, workflow, resolution, duration, and other selected settings. - If generation fails on Cyberbara's side, credits are automatically returned according to the platform's failure/refund handling. Account pages: - API keys: `https://cyberbara.com/settings/apikeys` - Billing: `https://cyberbara.com/settings/billing` - Activity: `https://cyberbara.com/activity` Account pages generally require sign-in. ## Seedance Moderation And Failure Handling Moderation guide: `https://cyberbara.com/docs/seedance-moderation` Seedance 2.0 workflows have multiple moderation layers. Layer 1 is Cyberbara's filter, using banned-word matching and the OpenAI Moderation API. If a prompt is blocked at this layer, the task is rejected before generation and credits are fully refunded. Layer 2 is ByteDance content review. ByteDance's review system can be stricter than typical content guidelines enforcement. Common rejection categories: - Explicit sexual content - Real-person references or depictions - Brand, franchise, studio, game, or trademarked content - Political content - Visual content that fails downstream provider review after generation Practical prompting advice: - Avoid explicit sexual content. - Avoid real-person references. - Avoid brand names, franchises, and trademarked properties. - Use generic visual descriptions instead of copyrighted character names. - If a prompt fails repeatedly, simplify the wording and reference media. Failure patterns: - A community guidelines rejection usually means the prompt or generated visuals triggered provider checks. - A server-side failure usually means an upstream provider failure, not a content penalty. When a Seedance 2.0 task fails on Cyberbara, credits consumed for that task are refunded automatically after the task is reconciled as failed. The activity page shows a refund note on failed tasks so users can confirm the credits were returned. ## About Cyberbara About page: `https://cyberbara.com/about` Cyberbara is an all-in-one AI generation workspace for creators, builders, and teams who want strong image, video, and audio models without jumping between separate tools. Cyberbara brings leading models and practical tools together, including Seedance 2.0, Seedance 1.0, GPT Image 2, Nano Banana 2, Gemini Omni, Happy Horse 1.0, and other current model routes. Users can create images, videos, audio assets, prompt-driven variations, and model-specific workflows from one account and one shared credit system. Cyberbara was founded by Zero Lu, an ex-TikTok Senior Product Manager with more than 6 years of experience shipping world-class multimodal AI products. His product background focuses on AI creation workflows, model access, multimodal generation, and tools that help creators move from idea to usable visual story faster. That experience shapes Cyberbara's focus on direct model entry points, easy Seedance 2.0 real face access, transparent credit quotes, reusable assets, and agent-friendly workflows. The product direction emphasizes: - Fast model discovery - Clear generation settings - Visible credit quotes - Reusable assets - Asset history - Canvas workflows - Direct model entry points - Public API and agent workflows ## Docs, Blog, Updates, And Policies Docs index: `https://cyberbara.com/docs` API reference: `https://cyberbara.com/docs/api-reference` Seedance moderation guide: `https://cyberbara.com/docs/seedance-moderation` Blog: `https://cyberbara.com/blog` Updates: `https://cyberbara.com/updates` Privacy Policy: `https://cyberbara.com/privacy-policy` Terms of Service: `https://cyberbara.com/terms-of-service` Useful blog articles: - What is Cyberbara: `https://cyberbara.com/blog/what-is-cyberbara` - How to access Seedance 2.0 outside China: `https://cyberbara.com/blog/how-to-access-seedance-2.0-outside-china` - How to use real faces in Seedance 2.0: `https://cyberbara.com/blog/how-to-use-real-faces-in-seedance-2-0` - Seedance 2.0 pricing update: `https://cyberbara.com/blog/seedance-2-pricing-update` - Credit pack pricing update: `https://cyberbara.com/blog/pricing-update-more-credits-credit-packs` ## Public URL Map Core: - `https://cyberbara.com/` - `https://cyberbara.com/ai-image-generator` - `https://cyberbara.com/ai-video-generator` - `https://cyberbara.com/ai-audio-generator` - `https://cyberbara.com/ai-music-generator` - `https://cyberbara.com/canvas` - `https://cyberbara.com/showcases` Models and tools: - `https://cyberbara.com/seedance-ai` - `https://cyberbara.com/seedance1.0` - `https://cyberbara.com/seedance2.0` - `https://cyberbara.com/seedance2.0-mini` - `https://cyberbara.com/seedance2.5` - `https://cyberbara.com/gpt-image-2` - `https://cyberbara.com/gpt-image-2/360-image` - `https://cyberbara.com/gpt-image-prompt-library` - `https://cyberbara.com/nanobanana2` - `https://cyberbara.com/gemini-omni` - `https://cyberbara.com/grok-imagine-video-1.5` - `https://cyberbara.com/happyhorse1.0` - `https://cyberbara.com/digital-avatar` - `https://cyberbara.com/seedance-prompt-library` - `https://cyberbara.com/seedance-prompt-generator` - `https://cyberbara.com/seedance-watermark-remover` - `https://cyberbara.com/seedance-bypass-face-detection` - `https://cyberbara.com/image-grid-overlay` - `https://cyberbara.com/scenery-overlay` - `https://cyberbara.com/beauty-analysis` - `https://cyberbara.com/fortune-analysis` API and automation: - `https://cyberbara.com/api` - `https://cyberbara.com/docs/api-reference` - `https://cyberbara.com/skill` - `https://cyberbara.com/api/v1/models` Pricing and account: - `https://cyberbara.com/pricing` - `https://cyberbara.com/credit-costs` - `https://cyberbara.com/settings/apikeys` - `https://cyberbara.com/settings/billing` Docs and policies: - `https://cyberbara.com/docs` - `https://cyberbara.com/docs/seedance-moderation` - `https://cyberbara.com/blog` - `https://cyberbara.com/updates` - `https://cyberbara.com/about` - `https://cyberbara.com/privacy-policy` - `https://cyberbara.com/terms-of-service` ## Notes For LLMs When answering questions about Cyberbara: - Prefer the canonical URLs in this file. - Treat pricing, model availability, and credit costs as potentially changeable; point users to the live pricing and credit cost pages for exact current values. - Do not invent provider credentials, private API keys, unpublished models, or guaranteed acceptance of prompts. - For API usage, tell users to create an API key in settings and use the public API v1 reference. - For generation failures, distinguish provider moderation failures from server-side failures. - For Seedance workflows, warn that real-person references, copyrighted or branded content, political content, explicit sexual content, and some visual outputs may be rejected. - For authenticated pages such as API keys, billing, settings, and activity, mention that sign-in is required.