AI & LLM
Token, cost & context tools for LLMs
OpenAI API Cost Calculator — GPT-5.5 Pricing (2026)
GPT-5.5 is billed at $5.00/M input tokens and $30.00/M output tokens (as of 2026-07-03). At the reference workload of 1,000 input + 500 output tokens per API call, the cost is $0.02 — $0.005 for input and $0.02 for output. At 1,000 calls per day, that is $20.00/day. Output tokens cost 6× more than input tokens, so trimming response length has a larger impact than compressing the prompt. Enter your own token counts and daily volume in the full calculator to get a precise monthly forecast.
Reference workload: 1,000 input + 500 output tokens
- Cost per call: $0.02 (input $0.005 + output $0.02)
- Per 1,000 calls: $20.00
- Input price: $5.00/M tokens | Output price: $30.00/M tokens
- Output costs 6× more than input — trim responses to reduce cost
- Rates as of 2026-07-03 (GPT-5.5)
Frequently asked questions
How much does the OpenAI API cost per 1,000 calls with GPT-5.5?
At 1000 input + 500 output tokens per call — a typical short chat turn — GPT-5.5 costs $0.02 per call and $20.00 per 1,000 calls. For a 10,000-call/day app, that is $200.00/day. Input is $5.00/M tokens; output is $30.00/M tokens. Rates as of 2026-07-03.
Why does output cost more than input in OpenAI pricing?
Generating tokens requires an autoregressive forward pass for every output token, while input is processed in a single parallel forward pass. That compute asymmetry is reflected in pricing: for GPT-5.5, output tokens cost 6× the input rate ($30.00/M vs $5.00/M). Reducing max_tokens or using concise system prompts can cut API costs significantly.