OpenAI o3 pricing
A reasoning model that spends tokens thinking before answering. The list price understates the real cost, because reasoning tokens bill as output and can dominate the total.
List prices checked
By OpenAI · Frontier tier
Token pricing
Per million tokens, from the official price list.
| Token type | Price per 1M | Notes |
|---|---|---|
| Input | $2.00 | Everything you send, including the system prompt |
| Output | $8.00 | Everything the model generates, reasoning tokens included |
| Cached input | $0.500 | Repeated prefixes served from cache |
| Blended (3:1) | $3.50 | One number for ranking, at three input tokens per output token |
Source: OpenAI pricing (May 2026). Published list prices only — committed-use discounts, regional variation, and per-request charges are not included. Not yet independently re-checked against the vendor’s page.
Capabilities
| Property | Value |
|---|---|
| Context window | 200K tokens |
| Max output | 100K tokens |
| Inputs accepted | text, image |
| Cost to fill the context once | $0.40 |
What it costs per month
Token prices applied to three workload shapes. Input-heavy and output-heavy workloads rank models very differently.
In-app chat assistant
A support or help assistant handling a few thousand conversations a month, with short prompts and short answers.
$80/mo
- 20M input tokens$40
- 5M output tokens$40
- If fully cached$50
RAG over documents
Retrieval-augmented answering where each request stuffs several retrieved passages into the prompt, so input dominates.
$720/mo
- 300M input tokens$600
- 15M output tokens$120
- If fully cached$270
Coding agent
An agentic loop that reads files and writes patches, generating heavy output and re-sending large context on every turn.
$1,600/mo
- 400M input tokens$800
- 100M output tokens$800
- If fully cached$1,000
OpenAI o3 vs other frontier models
Same capability tier, ranked by blended token price.
| Model | Input | Output | Context | Compare |
|---|---|---|---|---|
| Mistral Large | $2.00 | $6.00 | 131K | Head to head |
| Cohere Command A | $2.50 | $10.00 | 256K | Head to head |
| GPT-5 | $1.25 | $10.00 | 400K | Head to head |
| Gemini 2.5 Pro | $1.25 | $10.00 | 1M | Head to head |
| Claude Sonnet 5 | $3.00 | $15.00 | 200K | Head to head |
| Grok 4 | $3.00 | $15.00 | 256K | Head to head |
Questions about pricing
- How much does the OpenAI o3 API cost in 2026?
- $2.00 per million input tokens and $8.00 per million output tokens, from OpenAI's published pricing. The table above converts that into a monthly figure for three common workloads.
- What is the OpenAI o3 context window?
- 200K tokens, with up to 100K tokens of output per request. Filling the window costs whatever that many input tokens cost — at $2.00 per million, a single full-context request runs about $0.40.
- Does OpenAI o3 support prompt caching?
- Yes. Cached input is billed at $0.500 per million tokens, roughly 4× cheaper than uncached input. For an agent that resends the same system prompt and file context on every turn, that discount usually matters more than the headline rate does.
- Is OpenAI o3 cheaper than the alternatives?
- On a blended 3:1 input-to-output basis it works out at $3.50 per million tokens. The comparison table below ranks it against the other frontier models in the dataset, and the head-to-head pages model both against identical workloads.