Make every AI token go further.
Estimate tokens, compare API costs, and see which models can fit your prompt — without sending your text to a server.
What would this request cost?
Estimated standard text-token cost using the prompt above, your expected output size, and optional monthly request volume. See the full LLM API pricing comparison →
| Model | Input | Output | Per request | 30-day estimate |
|---|
Prices are informational estimates in USD per provider-published token rates and may exclude caching, batch discounts, tools, regional premiums, taxes, or special long-context rules. Always verify the linked provider pricing before making purchasing decisions.
Will your prompt fit?
Compare your estimated prompt plus expected output against published model context limits.
| Model | Context | Max output | Status | Estimated remaining |
|---|
Built to answer cost questions quickly.
Use a focused calculator for each search intent, then move between tools without re-entering provider pricing data.
Tokens ↔ Words
Convert token budgets into rough word counts for prompts, docs, and content planning.
API Cost Calculator
Model monthly spend from input, output, request volume, and provider pricing.
Context Window Checker
See which models can accept a long prompt, document, transcript, or codebase.
HTML Token Savings
Compare raw HTML against cleaned Markdown to estimate token and inference-cost savings.
1M Token Cost
How much one million tokens costs across OpenAI, Claude and Gemini — quick-answer table plus live calculator.
LLM API Pricing
Side-by-side rates for OpenAI, Claude and Gemini with a workload cost calculator — verified weekly.
OpenAI API Pricing
GPT-5.6 rates per 1M tokens, the 272K long-context surcharge, and monthly workload costs.
Claude API Pricing
Opus 5, Sonnet 5 and Haiku 4.5 rates per 1M tokens, plus the cancelled Sonnet price increase.
Gemini API Pricing
3.6 Flash and Flash-Lite rates per 1M tokens, and the promo window ending Dec 31, 2026.
OpenAI vs Claude
Same workload billed both ways: tier-by-tier rates, break-even volumes and the 272K long-context rule.
Claude vs Gemini
Identical workloads priced on both providers, and why the cheapest answer flips on Jan 1, 2027.
OpenAI vs Gemini
Where Google wins outright, and why the budget-tier verdict flips above 272K input tokens.
Tokens in 1,000 Words
Conversion table by content type — prose, code, JSON, HTML — plus what 1,000 words costs per model.
100K Tokens in Words
~75,000 words and 150 pages, the conversion by content type, and the chunk size that avoids the 272K surcharge.
What Is a Context Window?
Window sizes and output caps for all 8 tracked models, and what it costs to fill one end to end.
What Is an AI Token?
The sub-word unit behind every invoice: ~4 characters, 0.75 words, and 2x more for code, JSON and HTML.
1M Tokens in Words
~750,000 words and 1,500 pages — and the 5 of 8 models that accept it in one request.
Prompt Weight Analyzer
Find repeated instructions, markup and filler — 28–54% of input tokens on the samples tested.
Useful now, exact where it matters later.
This V1 deliberately uses a lightweight browser-side estimator so the site stays static, fast, and private. Exact token counts differ across model tokenizers, so the result is labeled as an estimate rather than pretending to be universal.
The next technical upgrade can add model-specific tokenizer packages while keeping the same no-backend architecture.
Token calculator basics
Is this an exact token count?
No. This V1 is a fast cross-model estimate. Each provider and model family can tokenize the same text differently. Exact model-specific tokenization is a planned upgrade.
Does GetMoreTokens upload my prompt?
No. The calculator in this static version runs in your browser. There is no form submission or backend endpoint for the text box.
Why are input and output token prices different?
LLM providers commonly charge separate rates for processing prompt/context tokens and generating output tokens. The calculator keeps them separate so workload estimates are more realistic.
Are API prices always current?
They change. This V1 stores pricing in one editable JavaScript data file and shows a verification date. Provider pricing pages remain the source of truth.