Chinese AI workflows for people and Agents

What do you want to create?

Start with the outcome. ChinaAPI helps you inspect the required input, compatible model path, cost and limits before carrying the task into an account.

One account

Contract pre-flight

Recoverable async jobs

Task-first routingAgent-readable

Goal

Requirements then

Contract

Result

The model is a replaceable executor. Your task, budget, and acceptance criteria stay intact.

Create a videoText, image, or reference-to-video

Generate or edit imagesProduct, campaign, and editing workflows

Build with text modelsReasoning, coding, extraction, and Agents

Generate or understand audioSpeech, transcription, and audio understanding

Featured project · MandarinSpark

Build an AI Chinese speaking tutor.

See how ASR, an LLM coach, deterministic pinyin, and TTS become one reproducible workflow. Explore the project boundary and the model verification needed before production.

Owned reference projectAgent-ready build pathDash-first fixtures

Explore MandarinSpark

MandarinSpark Chinese speaking tutor prototype

Learn Chinese projectReference project · production integration pending

Selected workflow

Animate an image

Prepare one authorized reference image and describe the motion. This contract is inspectable; runtime submission stays disabled.

Contract locked · runtime disabled

Illustrative reference image

MotionSlow orbital camera move, fabric flowing softlyFormatDuration, ratio, and audio await capability evidenceAcceptanceBottle shape and color remain consistent

Authorized referenceMotion descriptionAcceptance criteria

Inspect the executor path

Seedance fixture candidate

Routing-stable · capability-incomplete

Not a recommendation or availability claim

Verification pending

Contract evidence

task-first.v1.0.0 is pinned. Read the exact model ID, field map, and remaining activation gates.

Privacy pre-flight

Region, retention, training use, and deletion remain unresolved. Do not transfer a source image yet.

Pre-flight estimate

Not availableNo estimate is published without a live pricing snapshot.

Charge behaviorUnknownNetwork interruptionResume original task

Inspect the locked workflow

Read the Agent Recipe

Read-only pre-flight. Upload and generation are not enabled here.

Start from the job

Inputs, constraints, and acceptance criteria are visible before choosing a model.

Explore all workflows

Recover with evidence

Resume the original task before retrying. Ask before changing provider or creating another billable attempt.

Read integration guides

One fact model

HTML, Markdown, JSON, and Agent Recipes share one versioned workflow record.

Generate an API request

More resources, languages & business support

One account · one fact model

Infrastructure follows the task.

ChinaAPI still provides one account, unified billing, and OpenAI-compatible access where supported. Workflow facts are also published as HTML, Markdown, JSON, and Agent Recipe artifacts.

Discover on www

Understand what the task needs and which facts remain unresolved.

Execute in Console

Live quotes, account state, API keys, tasks, usage, and billing stay in Dash.

Recover with evidence

Resume the original task before retrying and ask before a new provider or billable attempt.

See existing integration guides →

Explore by need

Pages for common buying questions.

Developer quickstart guidesCreate a key and connect Cursor, Cline, LiteLLM, LobeChat, or n8n. Cursor configuration guideConnect Cursor to the OpenAI-compatible endpoint and verify a Chinese coding model. Cline configuration guideSet up the OpenAI Compatible provider in the VS Code agent. LiteLLM configuration guideRoute LiteLLM proxy or SDK calls with openai/<model> and a custom API base. LobeChat configuration guideUse ChinaAPI as LobeChat's OpenAI provider for Chinese LLM families. n8n configuration guideCreate OpenAI credentials with a custom base URL for AI workflow nodes. Talk to solutionsBusiness rollout support for concrete volume, model-evaluation, or supplier questions. Chinese AI model directoryBrowse LLM, image, and video model access pages from one hub. Chinese AI model rankingsCompare Chinese LLMs by context window, USD input price, and capabilities. Chinese AI API accessBusiness access overview for LLM, image, video, and production API workloads. Tool integration guidesSet up Cursor, Cline, LiteLLM, LobeChat, and n8n with ChinaAPI. Chinese LLM API providersCompare GLM, Qwen, DeepSeek, Kimi, and MiniMax for enterprise workloads. Chinese AI API alternativesEvaluate alternative routes for OpenAI, Claude, and video API workloads. Chinese AI model comparisonsCompare model families before committing production volume. AI agent referenceStructured reference for answer engines and research agents citing ChinaAPI. Agent ecosystem updatesExternal, source-labelled agent activity signals—not a ChinaAPI ranking. AI video API accessSeedance, Wan, Hailuo, Kling, and Qwen Image for creative teams. LLM API cost reductionHow to evaluate lower-cost model routing. GLM-5.2 API accessFor teams evaluating Chinese frontier LLM alternatives. Qwen API accessLLM API access for RAG, SaaS inference, and support automation. DeepSeek API accessEvaluate DeepSeek for coding, reasoning, and business LLM workflows. Qwen Image API accessImage generation workflows for products, ads, and creative teams. Chinese AI video modelsSeedance, Hailuo, Kling, and Wan for business video workflows. Seedance vs KlingCompare Chinese AI video models with business evaluation criteria. Claude API cost reductionEvaluate Chinese LLM alternatives for selected Claude workloads. Seedance model accessBusiness use cases for AI video generation. Seedance API accessBusiness video generation access for agencies and creative teams. Kling API accessEvaluate Kling alongside Seedance, Hailuo, and Wan. Hailuo API accessAI video access and model comparison for business workflows. AI video API for agenciesCampaign concepts, ad variants, and product video production. OpenAI API cost reductionEvaluate lower-cost Chinese model routing for selected workloads. Chinese AI video model comparisonCompare Seedance, Kling, Hailuo, and Wan for business use. FAQAvailability, pricing, pilots, and qualification.

Capability centers & free tools

Start from the capability, not the model name.

Image APIGeneration and editing models behind one endpoint. Video APIText, image, reference, and editing video workflows. Text-to-Speech APIVoices, delivery control, and audio output formats. Speech-to-Text APITranscription models and their language coverage. Audio APIAudio understanding and speech-to-speech building blocks. API cost calculatorEstimate monthly USD spend before you commit volume. Public pricing datasetInspect model pricing and the methodology behind the data. API request generatorCopy a working curl, Python, or JavaScript request. Seedance 2 agent recipeRunnable async video recipe with polling that survives interruption.

Languages

English 日本語 한국어 ไทย Tiếng Việt Русский Bahasa Indonesia Português do Brasil Español Français

Business rollout support

Have a real rollout to discuss?

Self-serve developers can register immediately. Talk to solutions only for existing AI spend, expected production volume, supplier qualification, or a concrete creative workflow.

[email protected]