OpenAI’s GPT-5.5 model page lists text-token rates of $5.00 per 1 million input tokens, $0.50 per 1 million cached input tokens, and $30.00 per 1 million output tokens for gpt-5.5. GPT-5.5 Pro is listed separately at $30.00 per 1 million input tokens and $180.00 per 1 million output tokens. The available pricing excerpt does not identify a selected processing tier or explicitly define the currency, so the figures below are reported as displayed—not labeled as Standard-tier or guaranteed USD rates.
GPT-5.5 API rates
| API model | Input per 1M tokens | Cached input per 1M tokens | Output per 1M tokens |
|---|---|---|---|
GPT-5.5 (gpt-5.5) |
$5.00 | $0.50 | $30.00 |
GPT-5.5 Pro (gpt-5.5-pro) |
$30.00 | No cached-input discount listed | $180.00 |
GPT-5.5 Pro is a separate offering; don’t use its rates to estimate standard GPT-5.5 usage. The available Pro pricing excerpt says it can be used through the Responses API, including Batch API, and does not list a cached-input discount.
Estimate your GPT-5.5 API bill
For a basic estimate, separate input into uncached and cached tokens so the same input is not counted twice:
Estimated token cost = (uncached input tokens ÷ 1,000,000 × $5) + (cached input tokens ÷ 1,000,000 × $0.50) + (billable output tokens ÷ 1,000,000 × $30)
Uncached and cached input are mutually exclusive portions of total input. The formula uses the rates as listed; the available pricing excerpt does not establish a specific processing-tier label for them.
For example, 1 million uncached input tokens plus 1 million billable output tokens totals $35 at the listed rates: $5 for input plus $30 for output. If 1 million input tokens qualify for the listed cached rate instead, that input portion is $0.50; add output and any other applicable usage separately.
Worked production example
Suppose a service makes 10,000 calls per day, each with 3,000 input tokens and 1,000 total billable output tokens, for 30 days. Assuming the listed rates, no cached input, and no long-context adjustment:
- Daily input: 30 million tokens × $5 per million = $150
- Daily output: 10 million tokens × $30 per million = $300
- Daily estimate: $450
- 30-day estimate: $13,500
This is an illustrative calculation, not a quote. The output assumption is total billable output, not merely visible answer text: OpenAI says reasoning tokens count toward output usage and are billed as output tokens, even though they are not visible in the answer. (help.openai.com) Replace the assumed token counts with measured usage. This estimate excludes cached input, long-context adjustments, regional-processing uplift, and other charges not represented by these token counts.
Pricing qualifications
Long prompts: OpenAI’s GPT-5.5 pricing excerpt says prompts with more than 272,000 input tokens are priced at 2× input and 1.5× output for the full session, for Standard, Batch, and Flex. The excerpt does not equate a “session” with one request. If the listed rates are the applicable starting rates, the multipliers imply $10 per million input tokens and $45 per million output tokens for that full session; those amounts are calculations from the listed rates, not separately quoted long-context price rows.
Cached input: The GPT-5.5 model page lists cached input at $0.50 per million tokens, compared with $5.00 for input at the other listed rate. Use the cached rate only for tokens actually billed as cached. The available GPT-5.5 Pro excerpt lists no cached-input discount.
Batch and Flex: The available material identifies Standard, Batch, and Flex in the long-context rule, but does not establish exact GPT-5.5 Batch or Flex rates or savings. Don’t include an assumed discount in a production budget without checking the current pricing terms for the processing mode you select.
Regional processing: The GPT-5.5 pricing excerpt reports a 10% uplift for regional-processing (data-residency) endpoints. Whether it applies depends on the endpoint and processing configuration.
Bottom line
Use the listed GPT-5.5 rates—$5 per million input tokens, $0.50 per million cached input tokens, and $30 per million output tokens—as the starting figures for a transparent estimate, while confirming the applicable processing tier and currency before committing a budget. Keep GPT-5.5 Pro separate at $30 input and $180 output per million tokens. For production planning, use measured uncached input, cached input, and total billable output, and account for the full-session long-context rule where it applies.