FAQ
Frequently asked questions.
Answers to the questions people ask most about GotoRamp: what it is, where it's available, which Chinese AI models it opens with, how billing works and how your data is handled.
Company
About GotoRamp
- What is GotoRamp?
- GotoRamp is an API service that gives businesses access to Chinese AI models: text models from the DeepSeek, Qwen, Kimi, GLM and Doubao Seed families, billed per token, and Seedance video, billed per second. It serves Southeast Asia and the Middle East, in the countries where that can be done lawfully.
- Who operates GotoRamp?
- GotoRamp LLC, a US limited liability company based in Mountain View, California. Invoices come from GotoRamp LLC in US dollars.
- Is GotoRamp available now?
- GotoRamp is in early access: no customer accounts are live yet, and the site collects early-access requests. API keys and the API endpoint open with your invitation; request early access to join.
- Is GotoRamp affiliated with the companies that make the models?
- No. GotoRamp is an independent company and isn't affiliated with or endorsed by DeepSeek, Alibaba Cloud, Moonshot AI, Zhipu AI or ByteDance. Model names belong to their owners.
- What support does GotoRamp offer?
- Email support in English and Mandarin on UTC+8 business days, with replies within one business day. Write to support@gotoramp.ai, or sales@gotoramp.ai about plans.
Where we operate
Availability and countries
- Which countries can use GotoRamp?
- Early access opens first in Singapore, Malaysia, the Philippines and Thailand. Indonesia, Vietnam, the United Arab Emirates and Saudi Arabia are planned after local registration. Any country not listed isn't offered yet, and API requests from countries we don't serve are refused.
- What has to happen before Indonesia, Vietnam, the UAE and Saudi Arabia open?
- Local registration. Indonesia requires registration as a foreign electronic system operator, Vietnam requires tax registration for foreign suppliers of digital services, and the UAE and Saudi Arabia require VAT registration as a non-resident supplier.
- Why isn't Seedance video offered in the United States, Canada, the UK, Australia or New Zealand?
- Because Seedance isn't offered in those countries through its provider's official channel, so GotoRamp doesn't offer it there either. See availability for the full list.
- Is GotoRamp available in mainland China?
- No. GotoRamp doesn't serve mainland China, or countries and regions under US sanctions.
Models
Models
- Which models will GotoRamp offer?
- Text models from the DeepSeek, Qwen, Kimi, GLM and Doubao Seed families, and four Seedance video models, all opening with early access. The models page describes each family.
- What are the four Seedance models?
seedance-2.5is the newest, for final renders;seedance-2.0is for production, with full multimodal references and native audio;seedance-2.0-fastis for quicker iteration; andseedance-2.0-miniis for lowest-cost drafts.- Do all text models support tool calling, JSON output and image input?
- No. Tool calling, JSON output, vision input and context length vary by model version, so check the console or
GET /v1/modelsbefore you rely on a feature. - How does GotoRamp source its models?
- We offer a model only once we hold written terms for it through a channel whose terms allow our use in each market. We don't use shared keys, pooled consumer accounts or unofficial endpoints.
Billing
Billing
- How does GotoRamp billing work?
- You top up a prepaid balance in US dollars and usage is deducted from it. There's no subscription, no seat fee and no monthly minimum.
- How are text models billed?
- Per token. Input and output tokens are billed at separate rates, and each model has its own rates.
- How is Seedance video billed?
- Per second of video returned, at the rate for that model and resolution.
- Am I charged for failed requests?
- No. Failed requests aren't billed.
- What plans are there, and where can I see the rates?
- There are three plans: Pay as you go; Volume, with a monthly commitment, lower rates, higher concurrency and payment by invoice and bank transfer; and Enterprise, with reserved capacity, a data processing agreement and custom terms. Rates are shared with early-access accounts; see pricing for how the plans work.
- Do GotoRamp invoices include tax?
- Where GotoRamp LLC is registered for tax in your country, we add it to the invoice. Invoices are issued by GotoRamp LLC in US dollars.
Trust
Data, safety and compliance
- Does GotoRamp use my data to train models?
- No. We don't use customer prompts, files or outputs to train models.
- How long does GotoRamp keep generated videos?
- Generated videos can be downloaded for 24 hours; after that they aren't kept. Copy each video to your own storage as soon as the job succeeds.
- How long does GotoRamp keep prompts and references?
- GotoRamp keeps prompts and references for 30 days, then deletes them, unless they're needed for an abuse investigation.
- Can I sign a data processing agreement?
- Yes, on the Enterprise plan, which includes a DPA and custom terms. Contact sales@gotoramp.ai to start.
- How do I report abuse or a security issue?
- Email abuse@gotoramp.ai about misuse or content made with GotoRamp, with task IDs if you have them. Send security issues to security@gotoramp.ai and privacy requests to privacy@gotoramp.ai.
Developers
Technical
- What is the GotoRamp API base URL?
https://api.gotoramp.ai/v1, which opens with your early-access invitation. Requests use a bearer-token API key, and each key can have its own spend limit.- Can I use my existing OpenAI SDK code?
- Yes. GotoRamp's text API is OpenAI-compatible, so you change the base URL and API key and use a GotoRamp model ID such as
deepseek-chat; the OpenAI SDK guide walks through it. - Does the text API support streaming?
- Yes. Set
"stream": trueonPOST /v1/chat/completionsto receive the reply as it's generated. - How does video generation work through the API?
- It's asynchronous:
POST /v1/video/generationsreturns a task ID, and you pollGET /v1/video/generations/{task_id}until the video is ready. Download it within 24 hours; the quickstart has examples.
Start with a small balance. Scale when it works.
We're onboarding early-access accounts in small batches. Tell us what you're building; we reply within one business day.