How accurate are the LLM API cost estimates?

Pricing is based on the 2026-05 public rate cards from OpenAI, Anthropic, Google, Mistral, DeepSeek, and xAI. Prices change — always verify with the provider before committing to a budget.

Does this send my text to any server?

No. All calculations happen in your browser. No text, tokens, or usage data is sent anywhere.

tools.voiddo.com › llm-cost

LLM API Cost Calculator

Q: What does the Pro export include?

The CSV export includes all 20+ models with input/output price-per-million-tokens, your calculated monthly cost, annual projection, and cost per 1k requests — useful for vendor comparison docs and budget proposals.

Enter your expected usage and see your estimated monthly spend across 20+ LLM APIs — GPT-4o, Claude, Gemini, DeepSeek, Grok, Mistral, and more. All calculations are client-side. No text sent anywhere.

Prompt tokens per request

avg input tokens (system prompt + user message)

Completion tokens per request

avg output tokens generated per response

Requests per day

total API calls across all users/jobs

monthly cost comparison (sorted cheapest first — click headers to sort)

vendor	model	monthly cost	annual	per 1k req	$/M in	$/M out	relative

Pro Export this table as CSV — model names, your costs, annual projections, per-request costs. Useful for budget docs and vendor comparison sheets.

Pricing based on published API rate cards as of 2026-05. Input/output token prices shown per 1M tokens. DeepSeek prices via api.deepseek.com. Llama via Groq public pricing. Prices change — verify with your provider before committing to a budget. tokcount counts tokens from actual text files; this tool estimates budget from expected usage.

faq

How accurate are the estimates?

Based on published API rate cards (2026-05), assuming uniform usage across 30 days. Real costs vary with caching, batching, tier discounts, and provider-specific free quotas. Always verify pricing before setting a production budget.

Does this send my usage data anywhere?

No. All math runs in your browser. Nothing is sent to any server.

What does "Pro $9 CSV export" include?

The CSV includes all 20+ models with input/output price-per-million-tokens, your calculated monthly cost, annual projection, cost-per-1k-requests, and the token inputs you used — useful for vendor comparison documents and budget proposals.

What is tools.voiddo Pro?

A one-time $9 payment that supports the studio building these tools. It unlocks CSV export here and gives you early access to new tools before they’re publicly listed. No subscription, perpetual license. learn more →

How do I count tokens in my actual prompts?

Use tokcount — paste or pipe your text and it counts tokens for GPT-4o, Claude, Gemini, and 60+ more models. Combine it with this calculator: measure actual token counts in tokcount, then project monthly cost here.