About GPT-5 nano
GPT-5 nano is the smallest and fastest member of the GPT-5 family, aimed squarely at summarisation, classification and other high-frequency, low-complexity work. Despite its size it keeps the family's large context window, image input, structured outputs and reasoning-effort controls, which makes it unusually capable for a bottom-tier model — long documents can be processed whole rather than chunked.
The trade-off is depth: tasks needing multi-step reasoning or nuanced judgement belong a tier up, with nano handling the volume around them. In OpenAI's lineup it plays the utility role: the model you reach for when the answer matters less than the throughput.
Nano is the cheapest entry in OpenAI's lineup, and cached input drops it close to free — a high-volume service with a shared system prompt pays mostly for output. That shape suits its intended jobs: classification labels and short summaries keep output tiny, so per-request costs round to almost nothing at list price. The risk is misapplication rather than price: a task that quietly needs reasoning produces confidently wrong answers cheaply. A useful pattern is nano as the front line with automatic escalation to mini on low confidence; the monthly-cost mode above shows what each split costs at your traffic level.
Frequently asked questions
How much does the GPT-5 nano API cost?
GPT-5 nano costs $0.05 per million input tokens and $0.4 per million output tokens, with cached input at $0.005 per million (90% off).
How much does a typical chat message cost with GPT-5 nano?
A short chat message (about 500 input and 250 output tokens) costs roughly $0.0001 with GPT-5 nano. At 1,000 such requests a day, that is about $3.75 per month.
Is GPT-5 nano cheap compared to similar models?
GPT-5 nano has the #1 cheapest input rate of the 29 models we track. On a mixed workload (10K input / 2K output), it costs $0.0013 per request and $0.0018 for Gemini 2.5 Flash-Lite.
What does GPT-5 nano cost per 1,000 tokens?
Per 1,000 tokens, GPT-5 nano costs $0.00005 for input and $0.00040 for output, or $0.00001 for cached input. Providers quote rates per million tokens ($0.05/M in, $0.4/M out here), so divide by a thousand for the per-1K figure older pricing pages used.
How large a prompt can GPT-5 nano take?
GPT-5 nano accepts up to 391K tokens of input context and can generate up to 125K tokens of output per request. At its input rate, filling the entire 391K-token window costs about $0.020 per request before any output.
Related
Compare all models at once ·Count tokens in your prompt
All OpenAI model costs: GPT-5.6 Sol · GPT-5.6 Terra · GPT-5.6 Luna · GPT-5.4 · GPT-5 · GPT-5 mini · GPT-4o · GPT-4o mini · o3 · o4-mini
Last updated 2026-09-03. Prices verified against OpenAI's official pricing page; see the methodology.