Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.tensormesh.ai/llms.txt

Use this file to discover all available pages before exploring further.

Managing Billing

The Billing page is your primary tool for managing financial health, payment methods, and historical usage. A valid account balance is required to provision GPU resources.

Setting up Billing

To manage your funds and payment details, navigate to Management → Billing. Add Account Balance — From the Overview tab, click the + Add Balance button to load credits onto your account.
Payment Methods — Switch to the Payment Methods tab to securely enter and store credit card information.
Transaction History — View all past charges and credit additions under the Transactions tab.
Deployment Requirement: You must have at least one valid card on file to deploy GPU resources. Ensure your payment method is added before attempting to provision any infrastructure.

Understanding Your Costs

The Billing Overview provides real-time visibility into your spending through two primary modules:

Current Balance

Displays your total available credit. Use this to ensure your deployments remain active without interruption.

Current Billing

Shows your accumulated costs for the current month and your next scheduled billing date.

Spending Overview

Track your consumption trends across different timeframes to better predict future costs: Today — Real-time spending for the current calendar day.
Last Week — Total expenditure over the previous seven days.
Last Month — A aggregate view of your spending for the previous full month.

Serverless Usage

The Serverless Usage tab on the Billing page gives you a detailed breakdown of token consumption and costs across your serverless model deployments.

Viewing Serverless Usage

Navigate to Management → Billing and select the Serverless Usage tab. Use the period selector to filter by month or view all-time usage.

Usage Summary

Four summary cards show your aggregate serverless consumption:

Input Tokens

Total number of input (prompt) tokens sent to serverless models.

Output Tokens

Total number of output (completion) tokens generated by serverless models.

Cached Tokens

Tokens served from Tensormesh’s KV cache instead of being recomputed.

Total Cost

Your total serverless spend for the selected period.

Serverless Token Pricing

Serverless models are billed per token with three price tiers:
Token TypeDescription
Input Token PriceCost per prompt/input token
Output Token PriceCost per completion/output token
Cached Token PriceReduced cost for tokens served from KV cache
See Serverless Inference for available models, per-model pricing, and quick start examples.