How can we help?

CentralHubAI is a low-cost AI inference and agentic-workflow platform. It gives users access to multiple AI model families through chat, API access, image generation, local coding agents, CodeForge, a browser-based cloud coding agent, and other AI tools from one account.

You can use CentralHubAI for everyday AI chat, document and image analysis, image generation and editing, API integrations, agentic coding, software-development workflows, and other agentic tasks. The available tools continue to expand over time.

The CentralHubAI interface is currently available in English, French, Spanish, and Portuguese. AI models can understand many additional languages depending on the selected model. CentralHubAI is available globally, and free users can register even where a local paid payment method is not available.

You can sign up using email and password, phone number and password, or Google. Email or phone verification is required depending on the signup method. Users must be at least 13 years old.

Yes. Every registered account has a free quota that resets monthly based on the account's signup cycle. Unused free quota expires when the next monthly quota is issued.

CentralHubAI offers weekly and monthly subscription options, and promotional pricing may be available. Prices can change, so visit https://centralhubai.com/subscriptions for the current plans and pricing. Plan-specific limits and benefits are available on the subscription details page.

Depending on the current plan, a subscription can include up to approximately $200 worth of usage across eligible models. The subscription uses one shared allocation, but different models consume that allocation at different rates, so cheaper models provide more usage than premium models. Check https://centralhubai.com/subscriptions/plan-details/lite for the current allocation and model usage rules.

CentralHubAI provides access to multiple model families, including current OpenAI GPT models, Anthropic Claude Opus and Sonnet models, DeepSeek, GLM, and others. The catalog can change as providers release, rename, restrict, or retire models. Check https://centralhubai.com/codeforge-console/pricing for the current model catalog and pricing.

Yes. You can use pay-as-you-go API credits, a subscription, or both. Pay-as-you-go pricing varies by model and is primarily based on input tokens, output tokens, and, where applicable, cache-read and cache-write tokens. Current prices are listed at https://centralhubai.com/codeforge-console/pricing.

No. Purchased API credits do not expire. The minimum top-up amount may change, so the current minimum is shown when purchasing credits.

If API-credit fallback is disabled, requests that require additional usage return an error instead of charging your credit balance. If you explicitly enable API-credit fallback in Settings under Usage, CentralHubAI can automatically continue using your API-credit balance after the subscription allocation is exhausted.

Yes. Usage updates in real time as you use the service. Subscription allocation is available at https://centralhubai.com/codeforge-console/usage/subscription and API-credit usage is available at https://centralhubai.com/codeforge-console/usage/api-credits. Per-model usage breakdowns, per-request cost details, historical usage views, and spending alerts are planned.

Yes. You can switch between available subscription periods, with the change applied immediately and the remaining value of the current term taken into account on a prorated basis. Subscriptions renew automatically unless you cancel renewal from Account Settings by opening your profile/avatar menu. Cancelling stops the next renewal but keeps your paid access active until the end of the current term.

Subscription payments are generally non-refundable. Cancelling a subscription disables future renewal rather than ending the already-paid term immediately.

CentralHubAI supports credit/debit cards and Mobile Money where available. Visa, Mastercard, and compatible prepaid or virtual cards can be used when accepted by the payment provider. Available payment methods depend on the user's country.

Mobile Money availability depends on the current payment-provider coverage. Supported markets currently include Benin, Burkina Faso, Cameroon, Congo-Brazzaville, the Democratic Republic of the Congo, Gabon, Ghana, Côte d'Ivoire, Kenya, Malawi, Mozambique, Nigeria, Rwanda, Senegal, Sierra Leone, Tanzania, Uganda, and Zambia. Supported networks vary by market and can include operators such as Airtel, MTN/MTN MoMo, Orange, M-Pesa, Moov, Wave, YAS, AT, Zamtel, and others. The payment options shown at checkout are the source of truth for your account and country.

Mobile Money uses the supported transaction currency for the selected market. Card payments in Europe are generally displayed in euros (€), while other international card payments are generally displayed in US dollars ($). API-credit purchases are denominated in USD, and when Mobile Money is used the local amount is converted using the applicable exchange rate at the time of payment.

Mobile Money payments are usually confirmed almost instantly, but delays can happen. If money was deducted and your subscription or credits were not activated, contact CentralHubAI support and include the transaction reference. For declined card payments, contact support first for guidance, then retry or use another payment method if needed.

Open Account Settings and go to Billing to view your subscription and credit-purchase history and download available receipts or invoices. The invoice issuer and available billing fields depend on the payment method. Card billing details, company information, taxes, and applicable VAT may be handled by Stripe or the relevant payment provider.

CentralHubAI provides a grace period after a failed renewal. If the payment is not completed during that period, the paid subscription is cancelled and paid-plan access ends. Your underlying free account remains available.

Yes. You can manually choose a model before sending a message and switch models during an existing conversation. The newly selected model receives the existing conversation context. Normal chat does not currently use an automatic model selector.

Yes. Your private conversation history is accessible from the sidebar on the home screen. You can search, rename, delete, and export individual conversations. Conversation sharing through public links is not currently available.

Yes. Individual conversations can be exported as Markdown or PDF. A full-account data export is not yet available.

Chat supports common files such as PDF, DOCX, XLSX, TXT, code files, and images. Free users can currently attach up to 2 files per input with a maximum of 5 MB per file. Paid users can attach up to 10 files per input with a maximum of 10 MB per file. Some files are retained with the conversation for context, while others may only need temporary processing.

Yes, when the selected model supports vision. If a model does not support image input, the image upload or request will fail and you should choose a vision-capable model.

You can generate or edit images directly from chat by selecting the image tool, or use the dedicated image interface at https://centralhubai.com/tools/image-tool. Supported aspect ratios and resolutions can be selected where available. Image generation uses dedicated image models and has its own limits or quota rather than consuming the normal text-model allocation.

Yes. Images generated through the dedicated image tool are stored in a full image library where they can be viewed, downloaded in full quality, or deleted. Multiple-image generation from a single prompt is not yet implemented.

CodeForge is CentralHubAI's coding toolset. It combines model/API access, coding-agent workflows, setup guides, usage management, and a browser-based cloud coding environment.

The Cloud Coding Agent is a browser-based agentic development environment. You can ask the agent to work on your code, run commands and tests, install dependencies, make commits, and push changes. It also provides access to a full web-based VS Code environment and command-line tools.

Yes. The Cloud Coding Agent supports both GitHub and GitLab, including public and private repositories. GitHub or GitLab controls which repositories you authorize. Linked access can be revoked at any time.

Each user has an isolated persistent Debian-based cloud environment where projects are stored. Individual agent sessions use Git worktrees to further isolate work on repositories. Users can install packages, access the internet, configure environment variables or secrets, and use normal development tooling.

Current resource limits depend on the plan. The Lite plan currently includes 10 GB of disk space, while CPU and RAM resources may be shared dynamically between users. Check the subscription details page for current limits because resources can change over time.

A local coding agent is a compatible agentic harness running on your own computer, such as OpenCode, Codex, Claude-based harnesses, Hermes, or similar tools. CentralHubAI recommends OpenCode as a popular free and open-source option. Depending on the harness, an agent can read and modify project files, run terminal commands, tests and builds, and perform Git operations.

CentralHubAI provides dedicated setup guides at https://centralhubai.com/codeforge-console/setup/opencode and https://centralhubai.com/codeforge-console/setup/codex. A Windows and Linux command-line setup utility can also configure supported harnesses, ask for your API key, retrieve the current model list automatically, and let you select the configuration you want. macOS support for the setup utility is planned.

Yes. In most compatible applications, configure the CentralHubAI OpenAI-compatible base URL, your CentralHubAI API key, and a supported model ID. The base URL is https://codeforge-oaic.centralhubai.com/v1. The current model IDs are available through the CodeForge setup and pricing information.

Open CodeForge Console from the main site menu, then choose API Keys, or go directly to https://centralhubai.com/codeforge-console/api-keys. You can create multiple named API keys and revoke or delete them at any time. API keys should only be stored in trusted applications or secure backend/local environments and should never be exposed publicly.

Yes. CentralHubAI supports OpenAI-compatible APIs, including /chat/completions and, where supported by the selected model/provider, capabilities such as /responses, embeddings, image generation, audio, tool/function calling, and related endpoints. Standard OpenAI-compatible SDKs can generally be used by changing the base URL and API key.

Yes. Streaming is supported, and some models are stream-only by design. If a selected model requires streaming, that behavior comes from the model or provider rather than being a CentralHubAI limitation.

Very high concurrency or heavy automated usage can trigger rate limits depending on fair-use protections and infrastructure load. A 401 error usually means you should verify that your API key is still valid. A 429 response indicates a rate limit, and the response headers include a Retry-After value that tells the client when to retry.

Yes. Eligible requests use your available subscription allocation first. API credits are used after the subscription allocation is exhausted only when you have explicitly enabled API-credit fallback in Settings.

No. CentralHubAI is an inference platform and does not use your chats, prompts, uploaded files, or source code to train its own AI models. To generate a response, the necessary content is transmitted to the selected underlying model provider, whose own data-handling terms may apply.

Your local workspace, project files, terminal activity, and agent traces remain on your device unless the harness sends specific context for inference. CentralHubAI does not independently save local-agent traces. Context sent to an external model provider is subject to that provider's policies.

When you link GitHub or GitLab to the Cloud Coding Agent, access tokens may be stored so the service can access the repositories you authorize. You can revoke linked access at any time.

Yes. You can enable app-based two-factor authentication using authenticator apps such as Google Authenticator. Recovery keys are provided so you can recover access if you lose your authenticator device.

Yes. You can permanently delete your account from Settings. Associated chats, API keys, and linked credentials are cleared or invalidated. Certain order or transaction records may be retained for 90 days or longer when required for accounting, fraud prevention, legal, or operational purposes. CentralHubAI does not store your full card details because payment processing is handled by external payment providers.

Immediately revoke the affected API key, change your password, review your security settings, and contact CentralHubAI support. If automated account recovery fails, support can assist after identity verification.

Visit https://centralhubai.com/helps. Support is available through the account ticket system, email, and WhatsApp, and users can attach screenshots or files to tickets and review ticket history. Support is available in English, French, Spanish, and Portuguese, and paid users receive higher priority.

Documentation is available at https://centralhubai.com/docs and continues to expand. For service issues, use https://centralhubai.com/helps; a dedicated status service is planned. Planned maintenance is communicated to users in advance by email, WhatsApp, or both depending on the situation.

Automation and commercial API usage are allowed when users respect rate limits and fair-use protections. Fraud, credential theft, malware, spam, attacks on other systems, illegal activity, or attempts to abuse the service can result in suspension or an immediate ban.

CentralHubAI does not claim ownership of your prompts, uploaded files, source code, or generated outputs, and commercial use is allowed. You are responsible for ensuring that your use of generated content does not violate third-party rights, especially for images, media, recognizable people, brands, copyrighted characters, or other protected material.

Yes. You can integrate the API into your own applications and serve your own customers, subject to applicable rate limits, fair-use protections, and acceptable-use rules. Accounts are currently individual rather than team-based, and you are responsible for securing any API keys you choose to share or deploy.

No. CentralHubAI works through a web browser and is optimized for phones, tablets, and desktops. It can also be installed as a Progressive Web App (PWA). There are currently no separate native Android, iOS, Windows, or macOS apps. On a phone you can instruct the Cloud Coding Agent to perform tasks, but direct code editing is better suited to a medium/large tablet, laptop, or desktop.