OpenAI API pricing
Updated
OpenAI lists standard per-token prices for 37 models. The cheapest is GPT-5 Nano at USD 0.05 input and USD 0.40 output per 1M tokens; the dearest is GPT-5.5 Pro at USD 30 / USD 180. It publishes cached-input prices for 32 of them, plus batch, flex / spot, fast and ultrafast tiers (no priority). Checked 2026-11-14.
- Models with a standard price
- 37
- Median input / output
- USD 2 / USD 14
- Cached-input prices
- 32 of 37
- Cheapest host for
- 27 of 27 shared models
Which models is OpenAI the cheapest provider for?
Of the 27 models that OpenAI and at least one other provider both price at the standard tier, OpenAI has the lowest price for 27, ranked on a blend of 3 input tokens to 1 output token (all of them tied with another provider at the same price; the maker's own API is listed first on a tie). Another 10 models have a standard price from OpenAI only.
| Model | OpenAI input / output | Next cheapest | Saving on 3:1 blend | Providers |
|---|---|---|---|---|
| GPT-5.5 | USD 5 / USD 30 | DigitalOcean USD 5 / USD 30 | tie | 6 |
| GPT-5.6 Luna | USD 0.20 / USD 1.20 | DigitalOcean USD 0.20 / USD 1.20 | tie | 6 |
| GPT-5.6 Sol (promo) | USD 4 / USD 20 | DigitalOcean USD 4 / USD 20 | tie | 6 |
| GPT-5.6 Terra | USD 2 / USD 12 | DigitalOcean USD 2 / USD 12 | tie | 6 |
| GPT-6 Astra | USD 10 / USD 50 | DigitalOcean USD 10 / USD 50 | tie | 6 |
| GPT-6 Luna | USD 0.10 / USD 0.50 | DigitalOcean USD 0.10 / USD 0.50 | tie | 6 |
| GPT-6 Sol | USD 2 / USD 10 | DigitalOcean USD 2 / USD 10 | tie | 6 |
| GPT-6.1 Sol | USD 2 / USD 10 | DigitalOcean USD 2 / USD 10 | tie | 6 |
| GPT-5.4 | USD 2.50 / USD 15 | DigitalOcean USD 2.50 / USD 15 | tie | 4 |
| GPT-6.1 Terra | USD 1 / USD 5 | DigitalOcean USD 1 / USD 5 | tie | 4 |
| GPT-4.1 (2025-04-14) | USD 2 / USD 8 | DigitalOcean USD 2 / USD 8 | tie | 3 |
| GPT-4o mini | USD 0.15 / USD 0.60 | DigitalOcean USD 0.15 / USD 0.60 | tie | 3 |
| GPT-5 (2025-08-07) | USD 1.25 / USD 10 | DigitalOcean USD 1.25 / USD 10 | tie | 3 |
| GPT-5 Mini | USD 0.25 / USD 2 | DigitalOcean USD 0.25 / USD 2 | tie | 3 |
| GPT-5 Nano | USD 0.05 / USD 0.40 | DigitalOcean USD 0.05 / USD 0.40 | tie | 3 |
Showing the 15 with the most competing providers; the full price list below marks every model page.
How does OpenAI price the most widely hosted models?
The 6 models OpenAI sells that the most other providers also sell, with OpenAI's rank among them on the 3:1 blend.
| Model | OpenAI | Rank | Cheapest |
|---|---|---|---|
| GPT-5.5 | USD 5 / USD 30 | 1 of 6 | OpenAI (this page) |
| GPT-5.6 Luna | USD 0.20 / USD 1.20 | 1 of 6 | OpenAI (this page) |
| GPT-5.6 Sol | USD 4 / USD 20 | 1 of 6 | OpenAI (this page) |
| GPT-5.6 Terra | USD 2 / USD 12 | 1 of 6 | OpenAI (this page) |
| GPT-6 Astra | USD 10 / USD 50 | 1 of 6 | OpenAI (this page) |
| GPT-6 Luna | USD 0.10 / USD 0.50 | 1 of 6 | OpenAI (this page) |
OpenAI price list
USD per 1M tokens: 107 prices for 37 models, one row per model with any other service tiers stacked under the standard price. Sources checked 2026-11-14 unless a row says otherwise. "vs maker" compares a standard price with the model maker's own API on the 3:1 blend; a minus means OpenAI is cheaper. OpenAI is the maker of every model here, so there is no maker price to compare with.
| Model | Tier | Input | Cached | Output | Context | vs maker | Source |
|---|---|---|---|---|---|---|---|
| GPT Chat Latest | standard | USD 5 | USD 0.50 | USD 30 | not stated | own model | source |
| GPT Rosalind Discovery | standard | USD 5 | USD 0.50 | USD 25 | not stated | own model | source |
| GPT Rosalind Research | standard | USD 5 | USD 0.50 | USD 25 | not stated | own model | source |
| GPT-4.1 (2025-04-14) | standard batch fast | USD 2 USD 1 USD 3.50 | USD 0.50 — USD 0.875 | USD 8 USD 4 USD 14 | not stated | own model | source |
| GPT-4.1 Mini | standard batch fast | USD 0.40 USD 0.20 USD 0.70 | USD 0.10 — USD 0.175 | USD 1.60 USD 0.80 USD 2.80 | not stated | own model | source |
| GPT-4o | standard batch fast | USD 2.50 USD 1.25 USD 4.25 | USD 1.25 — USD 2.125 | USD 10 USD 5 USD 17 | not stated | own model | source |
| GPT-4o mini | standard batch fast | USD 0.15 USD 0.075 USD 0.25 | USD 0.075 — USD 0.125 | USD 0.60 USD 0.30 USD 1 | not stated | own model | source |
| GPT-5 (2025-08-07) | standard batch flex / spot fast | USD 1.25 USD 0.625 USD 0.625 USD 2.50 | USD 0.125 USD 0.0625 USD 0.0625 USD 0.25 | USD 10 USD 5 USD 5 USD 20 | not stated | own model | source |
| GPT-5 Mini | standard batch flex / spot fast | USD 0.25 USD 0.125 USD 0.125 USD 0.45 | USD 0.025 USD 0.0125 USD 0.0125 USD 0.045 | USD 2 USD 1 USD 1 USD 3.60 | not stated | own model | source |
| GPT-5 Nano | standard batch flex / spot | USD 0.05 USD 0.025 USD 0.025 | USD 0.005 USD 0.0025 USD 0.0025 | USD 0.40 USD 0.20 USD 0.20 | not stated | own model | source |
| GPT-5 Pro | standard batch | USD 15 USD 7.50 | — — | USD 120 USD 60 | not stated | own model | source |
| GPT-5 Search API | standard | USD 1.25 | USD 0.125 | USD 10 | not stated | own model | source |
| GPT-5.1 | standard batch flex / spot fast | USD 1.25 USD 0.625 USD 0.625 USD 2.50 | USD 0.125 USD 0.0625 USD 0.0625 USD 0.25 | USD 10 USD 5 USD 5 USD 20 | not stated | own model | source |
| GPT-5.2 | standard batch flex / spot fast | USD 1.75 USD 0.875 USD 0.875 USD 3.50 | USD 0.175 USD 0.0875 USD 0.0875 USD 0.35 | USD 14 USD 7 USD 7 USD 28 | not stated | own model | source |
| GPT-5.2 Pro | standard batch | USD 21 USD 10.50 | — — | USD 168 USD 84 | not stated | own model | source |
| GPT-5.3 Codex | standard fast | USD 1.75 USD 3.50 | USD 0.175 USD 0.35 | USD 14 USD 28 | not stated | own model | source |
| GPT-5.4 | standard batch flex / spot fast | USD 2.50 USD 1.25 USD 1.25 USD 5 | USD 0.25 USD 0.13 USD 0.13 USD 0.50 | USD 15 USD 7.50 USD 7.50 USD 30 | not stated | own model | source |
| GPT-5.4 Mini | standard batch flex / spot fast | USD 0.75 USD 0.375 USD 0.375 USD 1.50 | USD 0.075 USD 0.0375 USD 0.0375 USD 0.15 | USD 4.50 USD 2.25 USD 2.25 USD 9 | not stated | own model | source |
| GPT-5.4 Nano | standard batch flex / spot | USD 0.20 USD 0.10 USD 0.10 | USD 0.02 USD 0.01 USD 0.01 | USD 1.25 USD 0.625 USD 0.625 | not stated | own model | source |
| GPT-5.4 Pro | standard batch flex / spot | USD 30 USD 15 USD 15 | — — — | USD 180 USD 90 USD 90 | not stated | own model | source |
| GPT-5.5 | standard batch flex / spot fast | USD 5 USD 2.50 USD 2.50 USD 12.50 | USD 0.50 USD 0.25 USD 0.25 USD 1.25 | USD 30 USD 15 USD 15 USD 75 | not stated | own model | source |
| GPT-5.5 Cyber Preview | standard | USD 12.50 | USD 1.25 | USD 75 | not stated | own model | source |
| GPT-5.5 Pro | standard batch flex / spot | USD 30 USD 15 USD 15 | — — — | USD 180 USD 90 USD 90 | not stated | own model | source |
| GPT-5.6 Cyber | standard | USD 12.50 | USD 1.25 | USD 75 | not stated | own model | source |
| GPT-5.6 Luna | standard batch flex / spot fast | USD 0.20 USD 0.10 USD 0.10 USD 0.40 | USD 0.02 USD 0.01 USD 0.01 USD 0.04 | USD 1.20 USD 0.60 USD 0.60 USD 2.40 | not stated | own model | source |
| GPT-5.6 Solpromo | standard batch flex / spot fast | USD 4 USD 2 USD 2 USD 8 | USD 0.40 USD 0.20 USD 0.20 USD 0.80 | USD 20 USD 10 USD 10 USD 40 | not stated | own model | source |
| GPT-5.6 Terra | standard batch flex / spot fast | USD 2 USD 1 USD 1 USD 4 | USD 0.20 USD 0.10 USD 0.10 USD 0.40 | USD 12 USD 6 USD 6 USD 24 | not stated | own model | source |
| GPT-6 Astra | standard batch flex / spot fast ultrafast | USD 10 USD 5 USD 5 USD 20 USD 60 | USD 1 USD 0.50 USD 0.50 USD 2 USD 6 | USD 50 USD 25 USD 25 USD 100 USD 300 | not stated | own model | source |
| GPT-6 Astra for Lawpromo | standard | USD 10 | USD 1 | USD 50 | not stated | own model | source |
| GPT-6 Cyber | standard | USD 20 | USD 2 | USD 100 | not stated | own model | source |
| GPT-6 Luna | standard batch flex / spot fast | USD 0.10 USD 0.05 USD 0.05 USD 0.20 | USD 0.01 USD 0.005 USD 0.005 USD 0.02 | USD 0.50 USD 0.25 USD 0.25 USD 1 | not stated | own model | source |
| GPT-6 Rosalind | standard | USD 15 | USD 1.50 | USD 75 | not stated | own model | source |
| GPT-6 Sol | standard batch flex / spot fast | USD 2 USD 1 USD 1 USD 4 | USD 0.20 USD 0.10 USD 0.10 USD 0.40 | USD 10 USD 5 USD 5 USD 20 | not stated | own model | source |
| GPT-6.1 Sol | standard batch flex / spot fast ultrafast | USD 2 USD 1 USD 1 USD 4 USD 12 | USD 0.10 USD 0.05 USD 0.05 USD 0.20 USD 0.60 | USD 10 USD 5 USD 5 USD 20 USD 60 | not stated | own model | source |
| GPT-6.1 Terra | standard batch flex / spot fast | USD 1 USD 0.50 USD 0.50 USD 2 | USD 0.05 USD 0.025 USD 0.025 USD 0.10 | USD 5 USD 2.50 USD 2.50 USD 10 | not stated | own model | source |
| o3 (2025-04-16) | standard batch flex / spot fast | USD 2 USD 1 USD 1 USD 3.50 | USD 0.50 — USD 0.25 USD 0.875 | USD 8 USD 4 USD 4 USD 14 | not stated | own model | source |
| o3 Pro | standard batch | USD 20 USD 10 | — — | USD 80 USD 40 | not stated | own model | source |
- Cached "—" means OpenAI's page shows no cache price for that row; budget cached tokens at the input price.
- Context "not stated": OpenAI's pricing page does not give context windows.
- Promo: OpenAI's page describes 2 prices as promotional or discounted; they can end without notice. Each model page has the list price where the source gives one.
Report a wrong price or number (email to info@broadnet.ai, with this page's link filled in).
What service tiers does OpenAI publish?
- batch (27 rows). Across 27 models with both prices, the median batch price is 50% below standard. OpenAI row note: “Batch API (asynchronous).”
- flex / spot (20 rows). Across 20 models with both prices, the median flex / spot price is 50% below standard. OpenAI row note: “Flex processing.”
- fast (21 rows). Across 21 models with both prices, the median fast price is 2× standard (range 1.67× standard to 2.5× standard). OpenAI row note: “Fast mode (formerly Priority processing).”
- ultrafast (2 rows). Across 2 models with both prices, the median ultrafast price is 6× standard. OpenAI row note: “Short context (up to 272K input tokens); longer prompts bill at the long-context rate. Ultrafast mode.”
Cached input: 32 of 37 standard prices include a cache-read price, a median 10% of the input price. What cached input means.
What we recorded about OpenAI's pricing page: Prices from the Markdown version of the pricing page (pricing.md), which holds the same tables as the rendered page including the tier tabs. Regional processing (data residency) endpoints add 10% for models released on or after March 5, 2026. Realtime, audio, image, embedding and moderation models left out (not text/chat LLMs).
BroadNet Inference and OpenAI
BroadRouter is BroadNet's own site, so BroadNet is never ranked on this page. BroadNet Inference serves none of the models OpenAI prices. BroadNet Inference pricing.
More on OpenAI
- OpenAI's own pricing page (the source; checked 2026-11-14)
- OpenAI API alternative: open-model providers compared
- LLM API free tiers and rate limits: OpenAI's free tier and default limits next to other providers'
- All 41 providers compared · Cheapest LLM APIs right now · Download every price (CSV, JSON)
Quick answers
Is OpenAI OpenAI-compatible?
Yes. It is the reference API that the official OpenAI SDKs are built for; they use https://api.openai.com/v1 when you set neither baseURL nor OPENAI_BASE_URL (OpenAI Node SDK source, client.ts, checked 2026-11-15).
What is OpenAI's API base URL?
https://api.openai.com/v1, as OpenAI documents it. It is the SDK default, and the OPENAI_BASE_URL environment variable overrides it (OpenAI Node SDK source, client.ts, checked 2026-11-15).
Does OpenAI keep API prompts and outputs?
By default, for up to 30 days. OpenAI says abuse monitoring logs are generated for all API feature usage and "retained for up to 30 days" by default. Files, conversations and threads are kept until you delete them. Zero data retention covers eligible endpoints only and is "subject to prior approval by OpenAI" (OpenAI, data controls in the OpenAI platform, checked 2026-11-14).
Does OpenAI train on API data?
Not by default. OpenAI says data sent to the API "is not used to train or improve OpenAI models" unless you explicitly opt in (OpenAI, data controls in the OpenAI platform, checked 2026-11-14).
Does OpenAI have a free tier?
The lowest usage tier is named Free: "User must be in an allowed geography", with a usage limit of 100 USD a month. The page publishes no per-model limits for it and no free credit amount. Default limits (Build (first paid tier: 5 USD in total credit purchases), Luna): 5,000 requests a minute and 2,000,000 tokens a minute. Build tier. Astra, Sol and Terra: 5,000 RPM and 1,000,000 TPM. No requests-per-day limits listed (OpenAI rate limits guide, checked 2026-11-14).
Does OpenAI have a batch API?
Yes. OpenAI's Batch API guide gives a 50% discount and a 24-hour completion window, with up to 50,000 requests or 200 MB per batch (OpenAI Batch API guide, checked 2026-11-15).
Questions
How much does the OpenAI API cost per token?
OpenAI's standard prices run from USD 0.05 input and USD 0.40 output per 1M tokens (GPT-5 Nano) to USD 30 / USD 180 (GPT-5.5 Pro). The median across 37 standard prices is USD 2 input and USD 14 output per 1M, which is USD 0.000002 per input token. Input, cached input and output are billed separately.
What is the cheapest model on OpenAI?
GPT-5 Nano, at USD 0.05 input and USD 0.40 output per 1M tokens (USD 0.1375 on a 3:1 input-to-output blend).
Is OpenAI cheaper than other providers?
It depends on the model. Of 27 models that other providers also price, OpenAI is the cheapest standard price for 27 (all of them tied with another provider), for example GPT-5.5, GPT-5.6 Luna and GPT-5.6 Sol; another provider is cheaper for none; and it is the dearest for none. Its median rank is 1 of a median 3 providers, on a 3:1 input-to-output blend.
Does OpenAI have batch, cached or priority pricing?
Cached input: yes, for 32 of 37 standard prices, at a median 10% of the input price. Other tiers: batch on 27 rows (median input 50% below standard); flex / spot on 20 rows (median input 50% below standard); fast on 21 rows (median input 2× standard); ultrafast on 2 rows (median input 6× standard).
How much does a chatbot cost per month on OpenAI?
For 2,000 requests a day, 1,500 input tokens (60% cached), 300 output, over 30 days: USD 9.27 on GPT-5 Nano, USD 324.45 on GPT-5.3 Codex (the middle of OpenAI's list) and USD 5,940 on GPT-5.5 Pro. Try your own numbers.