# Atlas Cloud Build text and media AI workflows with one model-access account and API key. - Findings24: - Site: - Preços: Paid - Categorias: Ferramentas Dev IA, API de LLM ## O que é Atlas Cloud? Atlas Cloud is a developer platform for accessing language, image, video and other AI models through a shared account and API key. Choose a supported model and supply its required inputs to build chat or media-generation workflows without managing separate provider accounts. Chat completions can return directly or stream text. Image and video generation use separate asynchronous endpoints: submit a task, receive a prediction ID, then retrieve its status and output URLs. Inference uses a funded balance, while the Coding Plan has a separate quota and purchase flow. ## O que Atlas Cloud pode fazer? ### Multi-model access Select supported language and media models through Atlas Cloud, using the model-specific options and billing rates. ### LLM chat integration Use the OpenAI-compatible chat-completions endpoint for supported LLMs, with direct or streaming responses. This compatibility does not make every media API use the same request format. ### Asynchronous image and video tasks Submit media-generation requests and track them by prediction ID. Completed tasks return output URLs for your application to use. ### Media uploads Upload files through the media-upload endpoint when a generation workflow needs reference content, then use the selected model's required parameters. ### Shared inference billing Manage model usage through one account balance and review usage and costs in the console. Different models have different charge units. ## Casos de uso ### Build a chat application Connect a supported LLM through chat completions and process its text response or streaming chunks. ### Create a media pipeline Submit image or video jobs, poll their status and collect completed outputs instead of treating generation as an immediate chat response. ### Compare model options Test the same workload with available models and compare the resulting output, required inputs and billed usage before choosing an integration. ## Perguntas frequentes ### How is ordinary API inference priced? It is pay-as-you-go in USD from a funded account balance; a subscription is not required for these API calls. LLMs are token-billed, while media charges depend on the model, operation, resolution, duration or token usage. Use the selected model's current rate and options to estimate cost. ### Is every endpoint OpenAI-compatible? The LLM chat-completions endpoint uses that protocol. Image and video tasks use Atlas Cloud's separate generation and prediction endpoints, with model-specific inputs. One API key does not mean identical request schemas or features across all models. ### How do account credits work? Ordinary API charges use subscription credits first, then cash balance, then bonus credits. Purchased credits expire 365 days after purchase; promotions and redeem codes can have their own expiry. Insufficient available balance causes new API requests to be rejected. ### Is the Coding Plan the same as the inference balance? No. It uses coding-plan API keys and a separate quota. Monthly subscriptions last 30 days with monthly and weekly caps; the separate pay-as-you-go pack lasts 90 days. Coding-plan requests do not consume the normal account balance, and new purchases or renewals use Stripe rather than account or promotional credits. ## Informações rápidas - **Primary users**: Developers building model-backed applications - **Modalities**: Language, image, video and other supported media models - **Chat workflow**: Synchronous or streaming completions - **Media workflow**: Asynchronous task, prediction status and output URLs - **Billing**: Paid on-demand inference; separate Coding Plan quota