Last verified
FRONTIER LABFOUNDED 2015PUBLIC-BENEFIT CORPMICROSOFT-BACKEDCHATGPT MAKERS

OpenAI API Pricing

The lab behind ChatGPT and the GPT family. Founded as a nonprofit in 2015 by Sam Altman, Elon Musk, Ilya Sutskever, Greg Brockman and others; restructured into a capped-profit LLC in 2019 and finally into a public-benefit corporation on October 28, 2025. Ships the largest commercial model lineup, the largest consumer AI product, and — after the April 2026 funding round — the highest-valued private AI company in history at ~$852B.

Production models
8
+2 embeddings · +chat-latest alias
Founded
2015
10 years operating
SWE-bench leader
88.7%
GPT-5.5 (OpenAI-reported)
Cheapest tier
$0.20/M
GPT-5.4 nano input
Data retention
30days
default · ZDR opt-in
Last price move
27d ago
Apr 22 mass deprecation
§ 01 / LINEUP

The full roster.

Side-by-side →
FRONTIER · REASONING

GPT-5.5

OpenAI's current flagship coding/reasoning model.

Input
$5/M
Output
$30/M
1M ctx · vision
FRONTIER-PRO

GPT-5.5 Pro

OpenAI's higher-accuracy Pro tier for harder questions.

Input
$30/M
Output
$180/M
1.05M ctx · vision
BALANCED · MID-TIER

GPT-5.4

Affordable GPT tier at half the GPT-5.5 standard price.

Input
$3/M
Output
$15/M
1.05M ctx · vision
FRONTIER-PRO

GPT-5.4 Pro

Pro tier for GPT-5.4.

Input
$30/M
Output
$180/M
1.05M ctx · vision
FAST · LIGHTWEIGHT

GPT-5.4 mini

Smaller GPT-5.4 model for lower-latency coding, computer-use, and subagent workloads.

Input
$0.75/M
Output
$5/M
400K ctx · vision
BUDGET

GPT-5.4 nano

Cheapest GPT-5.4-class model for classification, extraction, ranking, routing, and small subagents.

Input
$0.20/M
Output
$1/M
400K ctx · vision
CODING-SPECIALIST

GPT-5.3-Codex

Specialized Codex model optimized for agentic coding tasks in Codex-like environments.

Input
$2/M
Output
$14/M
400K ctx · vision
EMBEDDING

text-embedding-3-large LAUNCHED JAN 2024

Most capable OpenAI embedding model.

Input
$0.13/M
Output
$0.00/M
input-only ctx · text-only
EMBEDDING

text-embedding-3-small LAUNCHED JAN 2024

Small OpenAI embedding model.

Input
$0.02/M
Output
$0.00/M
input-only ctx · text-only
CHAT

chat-latest LAUNCHED MAY 2026

Alias for the latest Instant model used in ChatGPT.

Input
$5/M
Output
$30/M
400K ctx · vision
DEPRECATED · 2026-04-22

GPT-4 LAUNCHED MAR 2023

Original GPT-4 (gpt-4-0613, formerly gpt-4-0314).

Input
$30/M
Output
$60/M
8K ctx · text-only
DEPRECATED · 2026-04-22

GPT-4 Turbo LAUNCHED NOV 2023

Latest snapshot was gpt-4-turbo-2024-04-09.

Input
$10/M
Output
$30/M
128K ctx · vision
DEPRECATED · 2026-04-22

GPT-4o LAUNCHED MAY 2024

Latest snapshot was gpt-4o-2024-05-13.

Input
$5/M
Output
$15/M
128K ctx · vision
DEPRECATED · 2026-04-22

o4-mini LAUNCHED APR 2025

Successor to o3-mini — same headline rate but cheaper cache hit pricing ($0.275 vs $0.55).

Input
$1/M
Output
$4/M
200K ctx · text-only
DEPRECATED · 2026-04-22

GPT-3.5 Turbo LAUNCHED MAR 2023

Latest snapshot was gpt-3.5-turbo-0125 priced at $0.50/$1.50.

Input
$0.50/M
Output
$2/M
16K ctx · text-only
DEPRECATED · 2026-04-22

o3-mini LAUNCHED JAN 2025

First budget-tier reasoning model — 14x cheaper than original o1 ($1.10 vs $15 input) at competitive reasoning quality.

Input
$1/M
Output
$4/M
200K ctx · text-only
DEPRECATED · 2026-04-22

o1 LAUNCHED DEC 2024

First production o-series reasoning model (o1-2024-12-17 snapshot).

Input
$15/M
Output
$60/M
200K ctx · text-only
LEGACY · STILL SUPPORTED

GPT-4.1 LAUNCHED APR 2025

OpenAI's GPT-4.1 series introduced 1M context windows to the GPT-4 family alongside another frontier price compression to $2/$8.

Input
$2/M
Output
$8/M
1M ctx · vision
LEGACY · STILL SUPPORTED

GPT-4.1 mini LAUNCHED APR 2025

Mid-tier GPT-4.1 — same 1M context as the base model, ~5x cheaper.

Input
$0.40/M
Output
$2/M
1M ctx · vision
LEGACY · STILL SUPPORTED

GPT-4o mini LAUNCHED JUL 2024

Set a new cheap-tier floor at OpenAI when it launched — $0.15/$0.60 was a 33x compression vs GPT-3.5 Turbo's original $5/$15-equivalent positioning and made OpenAI competitive with DeepSeek and Gem…

Input
$0.15/M
Output
$0.60/M
128K ctx · vision
RETIRED · JUN 2025

GPT-4 32K LAUNCHED MAR 2023

Extended-context GPT-4 variant.

Input
$60/M
Output
$120/M
32K ctx · text-only
RETIRED · JUL 2025

GPT-4.5 preview LAUNCHED FEB 2025

Short-lived experimental flagship from OpenAI — only available for ~5 months.

Input
$75/M
Output
$150/M
128K ctx · vision
RETIRED · DEC 2024

GPT-4 Vision Preview LAUNCHED NOV 2023

First public OpenAI model with image input (alongside gpt-4-turbo's 128K context).

Input
$10/M
Output
$30/M
128K ctx · vision
RETIRED · OCT 2025

o1-mini LAUNCHED SEP 2024

Mini reasoning model launched alongside o1 — 5x cheaper than full o1 and surprisingly competitive on coding benchmarks.

Input
$3/M
Output
$12/M
128K ctx · text-only
§ 02 / SHELF

All side-by-side.

Methodology →
Model Input /M Output /M Cached Context Max output Vision Tools Tier
GPT-5.5 FLAGSHIP $5.00 $30 $0.5−90% 1M Frontier
GPT-5.5 Pro $30 $180 1.05M Active
GPT-5.4 $2.50 $15 $0.25−90% 1.05M Mid
GPT-5.4 Pro $30 $180 1.05M Active
GPT-5.4 mini $0.75 $4.50 $0.07−90% 400K Light
GPT-5.4 nano $0.2 $1.25 $0.02−90% 400K Active
GPT-5.3-Codex $1.75 $14 $0.17−90% 400K Active
text-embedding-3-large $0.13 Free input-only Active
text-embedding-3-small $0.02 Free input-only Active
chat-latest $5.00 $30 $0.5−90% 400K Active
GPT-4 DEPRECATED $30 $60 8K Deprecated
GPT-4 Turbo DEPRECATED $10 $30 128K Deprecated
GPT-4o DEPRECATED $5.00 $15 $2.50−50% 128K Deprecated
o4-mini DEPRECATED $1.10 $4.40 $0.28−75% 200K Deprecated
GPT-3.5 Turbo DEPRECATED $0.5 $1.50 16K Deprecated
o3-mini DEPRECATED $1.10 $4.40 $0.55−50% 200K Deprecated
o1 DEPRECATED $15 $60 $7.50−50% 200K Deprecated
GPT-4.1 LEGACY $2.00 $8.00 $0.5−75% 1M Legacy
GPT-4.1 mini LEGACY $0.4 $1.60 $0.1−75% 1M Legacy
GPT-4o mini LEGACY $0.15 $0.6 $0.07−50% 128K Legacy
GPT-4 32K RETIRED Retired Jun 6, 2025 → gpt-4o Retired
GPT-4.5 preview RETIRED Retired Jul 14, 2025 → gpt-4-1 Retired
GPT-4 Vision Preview RETIRED Retired Dec 6, 2024 → gpt-4o Retired
o1-mini RETIRED Retired Oct 27, 2025 → o4-mini Retired

The last 12 months at OpenAI.

Major releases, deprecations, and corporate events. Sourced from official OpenAI announcements, the developers.openai.com deprecation queue, and recent press.

APR 23 · 2026
GPT-5.5 + GPT-5.5 Pro released — new flagship leads SWE-bench Verified at 88.7% (OpenAI-reported)
RELEASE
APR 22 · 2026
API deprecation wave — GPT-3.5 Turbo, GPT-4, GPT-4 Turbo, GPT-4o, o1, o3-mini, o4-mini queued for retirement on Oct 23, 2026
PRICING
APR · 2026
Funding round closes at $122B raised, valuing OpenAI at ~$852B
CORPORATE
MAR 17 · 2026
GPT-5.4 mini + nano released — light and budget tiers added to the 5.4 family
RELEASE
MAR 05 · 2026
GPT-5.4 + GPT-5.4 Pro released — mid-tier flagship between GPT-5.3-Codex and GPT-5.5
RELEASE
FEB 13 · 2026
ChatGPT retires GPT-4o, GPT-4.1, GPT-4.1 mini, and o4-mini consumer access (API unchanged at this date)
PRICING
FEB 05 · 2026
GPT-5.3-Codex released — coding-specialist model with low/medium/high/xhigh reasoning effort
RELEASE
FEB · 2026
Codex desktop app launched — agentic coding tool for the laptop
FEATURE
FEB · 2026
Amazon commits $50B, SoftBank and Nvidia $30B each to OpenAI's Series F
CORPORATE
OCT 28 · 2025
Corporate restructure: capped-profit LLC → public-benefit corporation; OpenAI Foundation holds 26%, Microsoft 27%
CORPORATE
OCT 21 · 2025
ChatGPT Atlas browser launched (macOS first) — Chromium fork with ChatGPT-native sidebar agent
FEATURE
SEP 23 · 2025
Five more Stargate sites announced — Texas, New Mexico, Ohio; ~7 GW planned capacity, >$400B over 3 years
CORPORATE
JAN 21 · 2025
Stargate Project announced with SoftBank, Oracle, MGX — up to $500B US AI infra by 2029
CORPORATE
§ 04 / ACCESS

Where to get it.

Methodology →
§ 04 / BEST FOR

Which OpenAI for what.

More scenarios →
If you need maximum accuracy on hard reasoning or coding…
GPT-5.5 Pro
Profile →
If you want a frontier daily driver with cache discount…
GPT-5.5
Profile →
If you need balanced cost vs quality for production traffic…
GPT-5.4
Profile →
If you're doing high-volume classification, routing, or extraction…
GPT-5.4 nano
Profile →
If you live in a terminal and want an agentic coding model…
GPT-5.3-Codex
Profile →
If you're building RAG / semantic search
text-embedding-3-large
Profile →
If you need EU data residency for compliance…
GPT-5.5 via Azure EU Data Zone
Azure docs →
§ 06 / BACKGROUND

The company behind it.

openai.com →

OpenAI was founded on December 11, 2015 in San Francisco as a nonprofit research lab. The original founders included Sam Altman, Elon Musk, Ilya Sutskever, Greg Brockman, and Trevor Blackwell, with an initial $1B funding pledge from a group of donors. The stated mission: ensure that artificial general intelligence benefits all of humanity.

The corporate structure has gone through three eras. From 2015–2019 the lab ran as a pure 501(c)(3) nonprofit. In 2019 OpenAI created a capped-profit subsidiary (OpenAI LP, profit returns capped at 100×) to raise capital — Microsoft committed $1B that year and would eventually invest more than $13B for a 27% stake. In November 2023 the nonprofit board briefly removed and reinstated Sam Altman within five days, surfacing tension between the safety-focused board and the commercial entity. On October 28, 2025 the restructure completed: OpenAI converted into a public-benefit corporation with the OpenAI Foundation holding 26%, Microsoft holding 27%, and employees/investors holding 47%.

The technical contributions sit at the centre of the modern LLM era. The lab published the GPT-1 / GPT-2 / GPT-3 / GPT-4 / GPT-5 series, launched ChatGPT on November 30, 2022 (fastest product to 100M users), shipped DALL-E, Whisper, Sora, and the o-series reasoning models. As of 2026 the lineup centres on GPT-5.5 + GPT-5.4 with the older o-series and GPT-4 family queued for full API retirement on Oct 23, 2026.

Funding scaled hard in 2025–2026. After a $157B valuation in October 2024 and a $300B round in early 2025, OpenAI closed a ~$122B Series F at a ~$852B valuation in April 2026, with Amazon committing $50B and SoftBank and Nvidia $30B each in February 2026. In parallel, the Stargate Project announced in January 2025 — a $500B AI infrastructure JV with SoftBank, Oracle, and MGX — has added 5 new US sites and ~7 GW of planned capacity through 2026.

Compared to Anthropic: larger, broader model surface (image gen, audio, video, embeddings), bigger consumer brand via ChatGPT, but weaker stated safety commitments (no published RSP-equivalent) and a longer history of leadership volatility. Compared to Google DeepMind: weaker integration with productivity suites and a smaller context window on the flagship.

§ 07 / COMPETITORS

Other frontier labs.

All providers →

Frequently asked.

Practical questions about OpenAI, the GPT lineup, and how to spend on it without surprises.

Q · 01 Which GPT model should I start with? +
Default to GPT-5.5 at $5/$30 per M with 90% off cached input. Step up to GPT-5.5 Pro only for the hardest reasoning / coding questions where the accuracy delta is worth paying 6× output. Step down to GPT-5.4 for cheaper production traffic, GPT-5.4 nano for classification at $0.20/$1.25, or GPT-5.3-Codex for agentic coding inside the terminal.
Q · 02 How safe is OpenAI relative to Anthropic? +
OpenAI publishes the Preparedness Framework and now operates as a public-benefit corporation, but it has no published RSP-equivalent with hard capability thresholds, and the November 2023 board crisis showed real friction between the safety side and the commercial entity. For procurement teams that need documented safety commitments, Anthropic is the more conservative pick. For shipping speed, OpenAI.
Q · 03 Does OpenAI offer EU data residency? +
Yes via two paths. Direct API: OpenAI ships data residency in Europe for Enterprise + Edu customers. Cloud: Azure OpenAI Service exposes EU Data Zones across multiple EU regions — Sweden Central, France Central, Switzerland North among them — with processing and storage contractually pinned in-region. 28 Azure regions total host Azure OpenAI.
Q · 04 What's the data retention policy? +
Default API retention is 30 days for abuse monitoring; logs are not used for training. Zero Data Retention (ZDR) is available for eligible Enterprise and Edu customers — prompts and completions are not stored at all. ChatGPT consumer tiers have their own controls (chat history can be turned off). This is more retention than Anthropic Pro's zero-day default, less than Google's logging-by-default.
Q · 05 When will GPT-6 launch? +
OpenAI has not announced a numbering jump. Cadence so far: GPT-5 family launched mid-2025; 5.3, 5.4, 5.5 shipped on roughly two-month intervals through 2026. A 6.0 major bump is unlikely before late 2026 / early 2027 based on this rhythm — though OpenAI compresses dot-versions much faster than majors.
Q · 06 Can I get a discount? +
Three stackable discounts on the API: Batch API −50% (24-hour async jobs), Flex −50% (best-effort latency), and cached input −90% on the GPT-5.x main rows. Pro rows (5.5 Pro, 5.4 Pro) do not list a cached-input discount. Enterprise + scale-tier contracts add committed-use discounts on top. Azure OpenAI has its own Azure-side reservations.
Q · 07 How does OpenAI compare to other frontier labs? +
Biggest surface area (image / audio / video / embeddings / reasoning all under one roof), biggest consumer brand, deepest cloud partnerships. Weaker on stated safety commitments than Anthropic, weaker on long-context than Google's 2M, more expensive than DeepSeek by 10–30×. See side-by-side comparisons.