Skip to content

4 Uncensored LLM model added, currently available for only pay-as-you-go users.See Models

FAQ

Short answers to common questions about Tokens: OpenAI and Anthropic compatibility, models, privacy, paying in BDT, balances, keys, receipts, account deletion and uptime.

Last updated 17 Nov 2026

On this page

Short answers to the questions people ask most about Tokens, with links to the full docs. If your question is about a specific error code, Troubleshooting is faster.

API and compatibility#

Is Tokens OpenAI-compatible?#

Yes. Set the base URL to https://tokens.bd/v1 and use your Tokens key as the API key. Chat Completions, the legacy Completions endpoint, the Responses API, Models and Embeddings (for embedding models only) are supported, streaming included. Image generation, audio, files, batches, assistants, fine-tuning and moderations are not. Sending an image to a model in a chat request works: see Image input. See Overview and Chat Completions.

Is Tokens Anthropic-compatible?#

Yes. The Messages API is at https://tokens.bd/v1/messages. Tools and SDKs built for Anthropic take the base URL https://tokens.bd (no /v1) because they add the path themselves. You can call any model in the catalog this way, not only Claude models, because the gateway translates the request. See Messages and Claude Code.

Can I call Tokens from a browser app?#

No. Tokens' responses carry no CORS headers, so browser fetch calls fail, and a key in front-end code is visible to anyone who opens dev tools. Call Tokens from your server or a CLI, and have the browser call your server. Node.js has a Next.js route handler example.

Models#

Which models can I use?#

The catalog is curated by our team and spans several providers, including Claude, GPT, Gemini, Grok, DeepSeek, Kimi, GLM, Qwen and MiniMax model families, and it changes as models are added or retired. /models has the full list with prices and capabilities. Your key can see the models available to it right now with GET https://tokens.bd/v1/models. Model IDs look like deepseek/deepseek-v4.1-flash.

Do all models support tool calling?#

No. Tool calling, vision and reasoning depend on the model. Check the model page in /models before you rely on a feature. See Tool calling.

Privacy#

Do you store my prompts?#

No. Tokens stores usage metadata only: model, token counts, cost, latency, timestamps, the key used and the request ID. Prompt and response content isn't stored, and logs redact prompts. We don't train on your data or sell it. Your prompts are sent to the upstream model provider that serves the request, and that provider's own data policy applies. Details are on /security.

Billing and payments#

Is there a free tier?#

Current plans and what each includes are listed on /pricing. To create an API key you need a verified email and either an active plan or a positive wallet balance.

How do I pay in BDT?#

Prices are shown in USD, BDT, with BDT as the default. Payment methods currently available: bKash, Nagad, Visa Card, Master Card, Rocket, AMEX, DBBL Nexus, Google Pay, Internet Banking, SSL. Your wallet is kept in USD. When you pay in BDT, the amount is converted at the exchange rate locked at checkout. The minimum top-up is $5 in dollars or ৳500 in taka. See Paying in BDT.

What happens when my balance runs out?#

Requests fail with 402 insufficient_credits and a link to /dashboard/billing. Before that point, when your balance can't cover a request's full max_tokens, the gateway lowers max_tokens to what the balance covers, so replies can come back shorter than you expect. A low-balance alert (default $5) and usage alerts at 50/75/90/100% are under Notifications in the dashboard. Subscriptions are paid per period and don't renew on their own, so renew before the period ends. See Plans and wallet.

How do I get a receipt?#

Every payment creates a receipt (numbered REC-XXXXXXXX) under /dashboard/billing. Open it and press Download PDF, or Print. To put your company name and BIN on it, fill them in under Org Settings on /dashboard/team. Refunds are handled by our team under the policy at /refund-policy.

API keys#

Can I limit what a key can do?#

Yes, when you create it. Each key can have a monthly spend cap in USD and a list of allowed models. These can't be edited after creation, so to change them, create a new key and revoke the old one. Rate limits (requests per minute and concurrent requests) apply to your whole account, not per key. See API keys and Rate limits.

How do I rotate a key?#

In /dashboard/keys, rotate the key to get a new secret and choose when the old secret stops: at once (do this if the key leaked), or after 1 hour or 24 hours, so your apps and agents keep working while you move them to the new secret. Until then both secrets work; after that the old one gets 401 invalid_api_key. See API keys.

Account and reliability#

How do I delete my account?#

Contact support and ask for deletion. It's done manually by our team. Revoke your API keys first if you want them disabled right away.

Is there an SLA?#

There's no published SLA. /status shows current status and 90-day uptime for the Inference API, Dashboard & Auth, Billing and upstream providers, measured by real probes. If a provider fails, the gateway automatically fails over to another source for the same model where one is available.

How do I get help?#

Open a ticket from /dashboard/support. Include the x-tokens-request-id header from the failing response, the model ID and the time. Never include your API key. See Support.

Was this page helpful?

Still stuck? Open a support ticket

Need help configuring your agent?

Test your connection with the connection tester, or create an API key.