Monthly cost
GPT-5 Nano
By OpenAI · $0.0500 in / $0.4000 out per 1M tokens
Compare against all other models →Cost at common scales
| Tier Pre-baked workload sizes — Hobby (100 req/day), Small (10k), Mid (100k), Enterprise (1M). For your own numbers, use the main calculator. | In / Out tokens | Cache Higher cache hit % reduces input cost — typical for RAG and chatbot traffic. Glossary → | Monthly tokens (in / out) | Monthly cost |
|---|---|---|---|---|
| Hobby (100 req/day) Side projects, personal tools — 1k input / 200 output tokens, no caching. | 1,000 / 200 | 0% | 3,000,000 / 600,000 | $0.39 |
| Small (10k req/day) Early-stage SaaS — 2k input / 500 output tokens, 10% cache hit. | 2,000 / 500 | 10% | 600,000,000 / 150,000,000 | $87.00 |
| Mid (100k req/day) Mid-market product — 2k input / 500 output tokens, 30% cache hit (typical for chat with system prompts). | 2,000 / 500 | 30% | 6,000,000,000 / 1,500,000,000 | $810.00 |
| Enterprise (1M req/day) High-volume B2B — 2k input / 500 output, 50% cache hit (heavy RAG-style traffic). | 2,000 / 500 | 50% | 60,000,000,000 / 15,000,000,000 | $7500.00 |