Estimate your GPT-4o, o1, GPT-4o mini, and DALL·E 3 costs instantly. Enter your token usage and request volume — get exact monthly and annual costs. No signup, no data sent anywhere.
Select your model, enter token counts and request volume — results update live.
All prices in USD per million tokens, pay-as-you-go, June 2026.
| Model | Input / 1M tokens | Output / 1M tokens | Context window | Best for |
|---|---|---|---|---|
| GPT-4o | $2.50 | $10.00 | 128K | General flagship tasks |
| GPT-4o miniBEST VALUE | $0.15 | $0.60 | 128K | High-volume, simple tasks |
| o1 | $15.00 | $60.00 | 200K | Complex reasoning |
| o1-mini | $3.00 | $12.00 | 128K | STEM reasoning, cheaper than o1 |
| GPT-4 Turbo | $10.00 | $30.00 | 128K | Legacy — migrate to GPT-4o |
50% off all models via the Batch API — requests processed asynchronously within 24h. Ideal for document processing, content generation, analytics. See the Batch row in the calculator above.
$0.040 / 1024px standard · $0.080 / 1024px HD · $0.120 / 1792px HD. Compare vs Stable Diffusion →
Match your use case to the cheapest model that meets your quality bar.
Classification, extraction, chat, summarization, simple Q&A. 16× cheaper than GPT-4o.
Complex analysis, code generation, nuanced writing, multimodal tasks.
Hard math, multi-step reasoning, research. Use sparingly — premium priced.
STEM reasoning at lower cost than o1. Good for coding and math.
GPT-4o: $2.50/M input, $10.00/M output. GPT-4o mini: $0.15/M input, $0.60/M output. o1: $15.00/M input, $60.00/M output. DALL·E 3 standard: $0.040/image. All prices are pay-as-you-go USD (June 2026).
Monthly cost = (avg input tokens × input price/1M + avg output tokens × output price/1M) × monthly requests. The calculator above runs this formula live. Log your actual token usage via the OpenAI usage dashboard to get accurate input values.
Yes — 16× cheaper on both input and output. GPT-4o mini costs $0.15/$0.60 per million tokens vs GPT-4o's $2.50/$10.00. For classification, extraction, chat, and summarization tasks, GPT-4o mini typically delivers comparable quality.
The Batch API processes requests asynchronously with up to 24h turnaround at exactly 50% of standard pricing. Same model quality, same API — just not real-time. Any workload that doesn't need immediate response (document processing, content generation, overnight analytics) qualifies.
Top 5 strategies: (1) Use GPT-4o mini for simpler tasks — 16× cheaper. (2) Use Batch API for async workloads — 50% discount. (3) Trim your system prompt — every saved token is billed on every request. (4) Set max_tokens explicitly to prevent runaway output costs. (5) Cache responses for repeated queries. Full 7-strategy guide →
Pricing changes, optimization strategies, new calculators. No spam.