About o3
o3 comes from OpenAI's o-series, the line that established extended reasoning as a product: models that spend hidden chain-of-thought tokens working through a problem before answering. In its generation, o3 was the deep-reasoning flagship for mathematics, science and hard coding problems.
The unified GPT-5 design later absorbed exactly that capability — adjustable reasoning effort inside a general model — and OpenAI lists o3 as succeeded by GPT-5. It remains available, with text-and-image input and text output, for integrations built against o-series behaviour, particularly those tuned around its deliberate response style.
Reasoning models bill their thinking as output tokens, so o3's effective cost on a hard problem was always dominated by tokens you never see — budget by reasoning depth, not prompt length. Its list rates sit above GPT-5's input rate despite the successor relationship, and its cached-input discount is shallower than the GPT-5 family's, so there is no pricing case for starting new work here. OpenAI's stated succession path is the practical guidance: new reasoning workloads begin on GPT-5 with effort set high, while o3 serves continuity for pipelines that standardised on it.
Frequently asked questions
How much does the o3 API cost?
o3 costs $2 per million input tokens and $8 per million output tokens, with cached input at $0.5 per million (75% off). Note: OpenAI lists o3 as succeeded by GPT-5.
How much does a typical chat message cost with o3?
A short chat message (about 500 input and 250 output tokens) costs roughly $0.0030 with o3. At 1,000 such requests a day, that is about $90.00 per month.
Is o3 cheap compared to similar models?
o3 has the #19 cheapest input rate of the 29 models we track. On a mixed workload (10K input / 2K output), it costs $0.036 per request, versus $0.030 for Mistral Medium 3.5 and $0.040 for Claude Sonnet 5.
What does o3 cost per 1,000 tokens?
Per 1,000 tokens, o3 costs $0.0020 for input and $0.0080 for output, or $0.00050 for cached input. Providers quote rates per million tokens ($2/M in, $8/M out here), so divide by a thousand for the per-1K figure older pricing pages used.
How large a prompt can o3 take?
o3 accepts up to 195K tokens of input context and can generate up to 98K tokens of output per request. At its input rate, filling the entire 195K-token window costs about $0.400 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-5 nano · GPT-4o · GPT-4o mini · o4-mini
Last updated 2026-09-03. Prices verified against OpenAI's official pricing page; see the methodology.