Frequently Asked Questions

Common questions about Tokenlio and quick answers.

Getting Started

Do I need accounts with OpenAI, Anthropic, or Google?

No. Tokenlio handles all upstream provider relationships. You only need a Tokenlio account.

What's the minimum top-up amount?

$10 USD minimum to cover payment processing fees.

How long does it take to get started?

Typically 5-10 minutes:

1. Sign up (1 min)
2. Verify email (1 min)
3. Top up balance (2 min)
4. Create API key (1 min)
5. Make first request (1 min)

Can I test before paying?

We don't offer a free tier, but $10 covers thousands of requests with cost-effective models like GPT-3.5 Turbo.

Billing & Pricing

How does billing work?

Tokenlio uses prepaid, pay-as-you-go pricing:

• Top up your balance (minimum $10)
• Pay per token consumed
• No subscriptions or monthly fees
• Top up again when needed

How are tokens calculated?

Tokens are counted by the model's tokenizer. Roughly:

• 1 token ≈ 4 characters
• 1 token ≈ ¾ of a word
• 100 tokens ≈ 75 words

Exact token counts are returned in each API response.

What are input and output tokens?

• Input tokens: Your prompt and conversation history
• Output tokens: The model's generated response

Output tokens typically cost more because they require generation.

Are failed requests charged?

No. Only successful completions consume your balance. Authentication errors, rate limits, and server errors are not charged.

Can I get a refund?

Yes, unused balance can be refunded within 30 days of top-up, minus consumed amount and processing fees. Email support@tokenlio.ai.

Do you offer volume discounts?

Not currently for standard accounts. Contact sales@tokenlio.ai for enterprise pricing.

What payment methods do you accept?

Major credit/debit cards via Stripe:

• Visa
• Mastercard
• American Express

What currency do you use?

All pricing is in USD. Your bank may apply conversion fees for non-USD cards.

API Keys & Authentication

How do I authenticate requests?

Include your API key in the Authorization header:

Authorization: Bearer YOUR_API_KEY

How many API keys can I create?

No hard limit. Create separate keys for different:

• Environments (dev, staging, production)
• Applications
• Team members (in organizations)

Can I see my API key again after creation?

No. The full key is shown only once at creation. If lost, create a new key.

How do I rotate API keys?

1. Create a new key
2. Update your application
3. Test the new key
4. Delete the old key

What if my API key is compromised?

1. Immediately disable the key in the console
2. Check usage logs for unauthorized activity
3. Create a new key
4. Update your application
5. Delete the compromised key

Do API keys expire?

No automatic expiration. Keys remain active until you disable or delete them.

Models & Capabilities

Which models are supported?

We support popular models from:

• OpenAI (GPT-4, GPT-3.5)
• Anthropic (Claude 3, Claude 3.5)
• Google (Gemini 1.5, Gemini 1.0)

See Models & Pricing for the complete list.

Can I request access to a specific model?

Yes. Email support@tokenlio.ai with your use case.

Do all models support streaming?

Yes, all chat models support streaming responses.

What about function calling?

Supported on:

• GPT-4, GPT-4 Turbo
• GPT-3.5 Turbo
• Claude 3+ models (via tool use)

Can I use vision models?

Vision capabilities (image understanding) are coming soon for:

• GPT-4 Vision
• Claude 3+ models
• Gemini 1.5 models

How do I choose the right model?

Start with GPT-3.5 Turbo for testing:

• Fast and cost-effective
• Good for most tasks

Upgrade to GPT-4 or Claude for:

• Complex reasoning
• Higher quality
• Specialized tasks

See our Model Selection Guide.

Usage & Rate Limits

What are the rate limits?

Default limits:

• 60 requests/minute (personal workspace)
• 120 requests/minute (organization workspace)

Contact support for higher limits.

How do I handle rate limits?

Implement exponential backoff:

import time

for attempt in range(3):
    try:
        response = client.chat.completions.create(...)
        break
    except RateLimitError:
        time.sleep(2 ** attempt)  # 1s, 2s, 4s

Can I view my usage history?

Yes, in the console:

• Usage: Token consumption and costs by time period
• Request Logs: Individual request details

How long is usage history kept?

• Request metadata: 30 days
• Billing transactions: Indefinitely
• Detailed logs: 30 days

Do you log my prompts and responses?

By default, we do NOT log:

• Prompt content
• Response content
• API keys
• Authorization headers

We only log technical metadata for billing and debugging.

Organizations & Teams

What's the difference between personal and organization workspaces?

Personal:

• Just you
• Your balance
• Your API keys

Organization:

• Multiple members
• Shared balance
• Individual API keys
• Centralized management

How do I create an organization?

1. Click your profile (top right)
2. Select Create Organization
3. Name your organization
4. Invite team members

Who can do what in an organization?

Owner:

• Manage all settings
• Top up balance
• View all usage
• Invite/remove members
• Manage all API keys

Member:

• Create their own API keys
• View their own usage
• Cannot see balance or billing

How do I invite team members?

1. Go to Organization > Members
2. Click Invite Member
3. Enter their email
4. They receive an invitation link
5. They must log in and accept

Can members see the organization balance?

No. Only the Owner can view balance and billing information.

What happens when I remove a member?

• Member loses access to organization workspace
• Their organization API keys are disabled
• Usage history remains (attributed to organization)
• Member keeps their personal workspace

Can I transfer organization ownership?

Not self-service currently. Contact support@tokenlio.ai for assistance.

Technical Questions

Is Tokenlio OpenAI-compatible?

Yes. Use the OpenAI SDK with a custom base URL:

from openai import OpenAI

client = OpenAI(
    api_key="YOUR_TOKENLIO_KEY",
    base_url="https://api.tokenlio.ai/v1"
)

What's the base URL?

https://api.tokenlio.ai/v1

Do you support webhooks?

Not currently. Usage and billing are available via the console and API.

Can I use Tokenlio with LangChain?

Yes:

from langchain.chat_models import ChatOpenAI

llm = ChatOpenAI(
    openai_api_key="YOUR_TOKENLIO_KEY",
    openai_api_base="https://api.tokenlio.ai/v1",
    model_name="gpt-4-turbo"
)

What about other frameworks?

Any framework that supports OpenAI's API will work with Tokenlio. Just change the base URL.

Is there a Python/Node.js SDK?

Use the official OpenAI SDK with custom base URL. No separate SDK needed.

Do you offer GraphQL?

No, REST API only.

Security & Privacy

How are API keys stored?

Keys are encrypted at rest and never logged. Full keys are shown only once at creation.

Do you store my prompts?

No. We don't store prompt content or model responses by default.

What data do you collect?

We collect:

• Account information (email, name)
• Billing transactions
• Usage metadata (tokens, model, timestamp)
• Technical logs for debugging

We do NOT collect:

• Prompt/response content
• API keys (after creation)
• Payment card numbers (handled by Stripe)

Is my data encrypted?

Yes:

• HTTPS/TLS for all API traffic
• Encrypted database storage
• Secure payment processing via Stripe

Do you train models on my data?

No. We don't use your data for model training. We're a gateway, not a model provider.

Are you GDPR compliant?

We follow data protection best practices. Contact privacy@tokenlio.ai for specific compliance questions.

Where are your servers located?

Infrastructure details available on request for compliance purposes.

Support

How do I get help?

• Check this FAQ
• Email support@tokenlio.ai
• Visit Contact page

What information should I include in support requests?

• Your account email
• Request ID (if API error)
• Error message
• Steps to reproduce
• Expected vs actual behavior

What are your support hours?

Email support is monitored Monday-Friday, 9 AM - 5 PM UTC. We aim to respond within 24 hours.

Do you offer phone support?

Not currently. Email is the primary support channel.

Is there a community forum?

Not yet. Currently support is via email.

Errors & Troubleshooting

I'm getting a 401 error

Cause: Invalid or missing API key

Solutions:

• Check your API key is correct
• Verify the key is active (not disabled)
• Ensure proper header format: Authorization: Bearer YOUR_KEY

I'm getting a 402 error

Cause: Insufficient balance

Solution: Top up your account balance.

I'm getting a 429 error

Cause: Rate limit exceeded

Solutions:

• Slow down request rate
• Implement exponential backoff
• Contact support for higher limits

My requests are slow

Possible causes:

• Using a slow model (GPT-4 is slower than GPT-3.5)
• Large output length
• Network latency

Solutions:

• Use faster models (GPT-3.5, Claude Haiku, Gemini Flash)
• Reduce max_tokens
• Use streaming for perceived speed

I'm not seeing my usage

Solutions:

• Check the correct workspace is selected
• Verify time range in usage filters
• Wait a few minutes for real-time sync
• Contact support if persists

Business Questions

Can I resell Tokenlio access?

Contact sales@tokenlio.ai to discuss partnership opportunities.

Do you offer white-label solutions?

Not currently. Contact sales@tokenlio.ai for custom requirements.

What's your SLA?

No formal SLA currently. We aim for >99% uptime. Enterprise SLAs available on request.

How do I report downtime?

Email support@tokenlio.ai with details.

Can I get a custom contract?

For enterprise customers, contact sales@tokenlio.ai.

Didn't Find Your Answer?

Contact us:

• Email: support@tokenlio.ai
• Visit: Contact page

We typically respond within 24 hours on business days.