AI API Cost Calculator
Compare and estimate LLM API costs across Claude, GPT, and Gemini by tokens and volume.
≈ 1,500 words
≈ 375 words
total calls to estimate
| Model | Input $/1M | Output $/1M | Cost / request | Monthly est. |
|---|---|---|---|---|
OpenAIGPT-4.1 nanoCheapest | $0.1 | $0.4 | $0.0004 | $4 |
GoogleGemini 2.5 Flash-Lite | $0.1 | $0.4 | $0.0004 | $4 |
GoogleGemini 2.5 Flash | $0.3 | $2.5 | $0.0019 | $18.5 |
AnthropicClaude Haiku 4.5 | $1 | $5 | $0.0045 | $45 |
OpenAIGPT-5 | $1.25 | $10 | $0.0075 | $75 |
GoogleGemini 2.5 Pro ≤200K context; higher above | $1.25 | $10 | $0.0075 | $75 |
OpenAIGPT-4o | $2.5 | $10 | $0.01 | $100 |
AnthropicClaude Sonnet 4.6 | $3 | $15 | $0.0135 | $135 |
AnthropicClaude Opus 4.8 | $5 | $25 | $0.0225 | $225 |
OpenAIGPT-5.5 | $5 | $30 | $0.025 | $250 |
AnthropicClaude Fable 5 | $10 | $50 | $0.045 | $450 |
Estimates only, standard pricing as of June 2026. Verify current prices on each provider's official pricing page before relying on these figures.
About the AI API cost calculator
Large language model APIs are priced per token — separately for input (your prompt) and output (the response). This calculator estimates and compares the cost of running the same workload across the major models from Anthropic (Claude), OpenAI (GPT), and Google (Gemini).
Enter the tokens per request and your monthly volume to see cost per call and projected monthly spend, ranked cheapest first. Use it to pick a model, budget a feature, or sanity-check an AI automation project.
Frequently asked questions
- How is the cost calculated?
- Cost per request = (input tokens ÷ 1,000,000 × input price) + (output tokens ÷ 1,000,000 × output price). Monthly cost multiplies that by your number of requests. All models use the same formula with their own per-million-token prices.
- How do tokens relate to words?
- Roughly 1 token ≈ 0.75 words (so ~1,000 words ≈ 1,300 tokens), but it varies by model and content. For exact counts use the provider's tokenizer or token-counting endpoint.
- Can I lower these costs?
- Yes — prompt caching and batch APIs typically cut input costs by 50–90% for repeated context, and smaller models (Haiku, GPT-4.1 nano, Gemini Flash-Lite) are far cheaper for simple tasks.
- Are the prices current?
- Prices shown are standard tier as of June 2026. Model pricing changes often — always confirm against each provider's official pricing page (linked below the table) before budgeting.