GPT-5 pricing
OpenAI's flagship, priced aggressively below the other frontier models on input and with a 90% cached-input discount that makes long system prompts nearly free to reuse.
List prices checked
By OpenAI · Frontier tier
Token pricing
Per million tokens, from the official price list.
| Token type | Price per 1M | Notes |
|---|---|---|
| Input | $1.25 | Everything you send, including the system prompt |
| Output | $10.00 | Everything the model generates, reasoning tokens included |
| Cached input | $0.125 | Repeated prefixes served from cache |
| Blended (3:1) | $3.44 | 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. Last checked against the vendor’s page on 2026-08-07.
Capabilities
| Property | Value |
|---|---|
| Context window | 400K tokens |
| Max output | 128K tokens |
| Inputs accepted | text, image |
| Cost to fill the context once | $0.50 |
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.
$75/mo
- 20M input tokens$25
- 5M output tokens$50
- If fully cached$53
RAG over documents
Retrieval-augmented answering where each request stuffs several retrieved passages into the prompt, so input dominates.
$525/mo
- 300M input tokens$375
- 15M output tokens$150
- If fully cached$188
Coding agent
An agentic loop that reads files and writes patches, generating heavy output and re-sending large context on every turn.
$1,500/mo
- 400M input tokens$500
- 100M output tokens$1,000
- If fully cached$1,050
GPT-5 vs other frontier models
Same capability tier, ranked by blended token price.
| Model | Input | Output | Context | Compare |
|---|---|---|---|---|
| Gemini 2.5 Pro | $1.25 | $10.00 | 1M | Head to head |
| OpenAI o3 | $2.00 | $8.00 | 200K | Head to head |
| Mistral Large | $2.00 | $6.00 | 131K | Head to head |
| DeepSeek-V4-Pro | $0.435 | $0.870 | 1M | Head to head |
| Cohere Command A | $2.50 | $10.00 | 256K | Head to head |
| Claude Sonnet 5 | $3.00 | $15.00 | 200K | Head to head |
Questions about pricing
- How much does the GPT-5 API cost in 2026?
- $1.25 per million input tokens and $10.00 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 GPT-5 context window?
- 400K tokens, with up to 128K tokens of output per request. Filling the window costs whatever that many input tokens cost — at $1.25 per million, a single full-context request runs about $0.50.
- Does GPT-5 support prompt caching?
- Yes. Cached input is billed at $0.125 per million tokens, roughly 10× 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 GPT-5 cheaper than the alternatives?
- On a blended 3:1 input-to-output basis it works out at $3.44 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.