Skip to content
inetGeek

OpenAI API pricing and limits

iScore46–64 · ≈#5/10
openai.com · LLM APIs
SEPT 2026 audit

Hosted inference for the GPT model family, sold per million tokens with separate short- and long-context rates.

Visit OpenAI documentation

No affiliate links, sponsored placements or paid rankings appear on this site. Ordering follows the sourced data and the stated criteria.

RegionsTen data-residency regions: US, Europe (EEA + Switzerland), Australia, Canada, Japan, India, Singapore, South Korea, United Kingdom, United Arab Emirates. Regional processing (inference) only in US, EU and UAE; the rest are storage-only.
CategoryLLM APIs

Reviewed by Palash BagchiUpdated 20 criteria comparedSources last checked

OpenAI API prices by model

API prices per modelStandard-processing prices in USD per 1M tokens as published on OpenAI's API pricing page, at the short-context rate. Notes gives the long-context rate, the cached-input price, and the Batch, Flex and Fast mode rates where the page lists them. Data-residency endpoints add 10% for eligible models released on or after March 5, 2026.
ModelInput (per 1M tokens)Output (per 1M tokens)Notes
gpt-6-astra$10.00$50.00Long context $20.00 in / $75.00 out. Cached input $1.00. Batch or Flex $5.00 / $25.00; Fast mode $20.00 / $100.00.
gpt-6-sol$2.00$10.00Long context $4.00 in / $15.00 out. Cached input $0.20. Batch or Flex $1.00 / $5.00; Fast mode $4.00 / $20.00.
gpt-6-luna$0.10$0.50Long context $0.20 in / $0.75 out. Cached input $0.01. Batch or Flex $0.05 / $0.25; Fast mode $0.20 / $1.00.
gpt-5.6-sol$4.00$20.00Cyber (Daybreak) model. Long context $8.00 in / $30.00 out. Cached input $0.40. Promotional pricing at least through November 21, 2026.
gpt-5.6-cyber$12.50$75.00Cyber (Daybreak) model. Cached input $1.25. No long-context price listed.
chat-latest$5.00$30.00ChatGPT model. Cached input $0.50.
gpt-5.3-codex$1.75$14.00Codex model. Cached input $0.175. Fast mode $3.50 / $28.00.

Pricing | OpenAI API · read · every row quoted from this page

01.Evaluation guide

Who it suits

Work that fits inside the short-context rate, where the cheapest tier is $0.20 per million input tokens and the batch tier halves everything above it.

Consider something else if…

Long prompts cross into the long-context rate and roughly double the bill, which is a cliff the flat-rate vendors do not have.

Editorial · Palash Bagchi · approved

Machine-derived signals

Not sourced facts — counts and matches read by a script, not a person. See each line for where it comes from.

OpenAI's own site (openai.com) self-detects as: hosting — Canadian hosting, Server Location, Cloud Hosting, Cloud PaaS, Indian hosting · CDN — Edge Delivery Network · DNS — Enterprise DNS, TLD Redirects (category-level via BuiltWith, not a named vendor)

Each value below links to the page it was read from, with the sentence it came from. Criteria OpenAI does not publish are not listed.

OpenAI API pricing

1 criteria
Pricing modelPer million tokens, with a second rate above the short-context threshold — long-context input and output cost roughly double. Batch, Flex, Fast and standard tiers are priced separately, and regional processing adds a 10% uplift on models released from March 2026.
Sources (1) →
  • Pricing — OpenAI API docs ↗

    “Prices per 1M tokens. [...] Regional processing (data residency) endpoints are charged a 10% uplift for models released on or after March 5, 2026 [...] Priority processing was renamed Fast mode on July 30, 2026.”

    Read 2026-09-06 · official pricing · Report a correction

Infrastructure

1 criteria
RegionsTen data-residency regions: US, Europe (EEA + Switzerland), Australia, Canada, Japan, India, Singapore, South Korea, United Kingdom, United Arab Emirates. Regional processing (inference) only in US, EU and UAE; the rest are storage-only.
Sources (1) →

Inference

8 criteria
Input price, top model$10 per million input tokens for gpt-6-astra at short context.
Sources (1) →
  • Pricing — OpenAI Platform ↗

    “Short context Long context Model Input Cached input Cache writes Output Input Cached input Cache writes Output gpt-6-astra $10.00 $1.00 $12.50 $50.00 $20.00 $2.00 $25.00 $75.00”

    Read 2026-09-14 · official pricing · Report a correction

Output price, top model$50 per million output tokens for gpt-6-astra at short context.
Sources (1) →
  • Pricing — OpenAI Platform ↗

    “Short context Long context Model Input Cached input Cache writes Output Input Cached input Cache writes Output gpt-6-astra $10.00 $1.00 $12.50 $50.00 $20.00 $2.00 $25.00 $75.00”

    Read 2026-09-14 · official pricing · Report a correction

Input price, cheapest model$0.10 per million input tokens for gpt-6-luna at short context.
Sources (1) →
Context window1,050,000 tokens on gpt-6-astra.
Sources (1) →
Max output tokens128,000 tokens on gpt-6-astra.
Sources (1) →
Prompt cachingSupported — Cached input is a tenth of the input rate; cache writes are priced above it.
Sources (1) →
Batch discount50% off input and output through the Batch API: gpt-6-astra falls from $10.00/$50.00 to $5.00/$25.00 at short context.
Sources (1) →
  • Pricing — OpenAI Platform ↗

    “All models Batch Short context Long context Model Input Cached input Cache writes Output Input Cached input Cache writes Output gpt-6-astra $5.00 $0.50 $6.25 $25.00 $10.00 $1.00 $12.50 $37.50”

    Read 2026-09-14 · official pricing · Report a correction

Rate limitsOrg- and project-level limits in RPM, RPD, TPM, TPD and IPM that vary by model; usage tiers graduate automatically as API spend rises.
Sources (1) →
  • Rate limits - OpenAI API docs ↗

    “Rate limits use metrics such as RPM (requests per minute), RPD (requests per day), TPM (tokens per minute), TPD (tokens per day), IPM (images per minute) [...] Rate limits are defined at the organization level and at the project level, not user level. [...] As your spend on our API goes up, we automatically graduate you to the next usage tier. This usually results in an increase in rate limits across most models.”

    Read 2026-09-13 · official docs · Report a correction

Compliance

10 criteria
SOC 2SOC 2 Type 2 report by an independent third-party auditor covering security, confidentiality, privacy and availability; access requested in the Trust Portal.
Sources (1) →
  • OpenAI Trust Portal ↗

    “Our products are covered in our SOC 2 Type 2 report and have been evaluated by an independent third-party auditor to confirm that our controls align with industry standards for security, confidentiality, privacy and availability.”

    Read 2026-09-13 · official site · Report a correction

ISO 27001ISO/IEC 27001:2022 certified, alongside 27017, 27018 and 27701; the certificate scope names the API, ChatGPT Enterprise and ChatGPT Edu.
Sources (1) →
  • OpenAI Trust Portal ↗

    “Our products are also ISO 27001, 27017, 27018, and 27701 certified. [...] for OpenAI's API, ChatGPT Enterprise, and ChatGPT Edu services.”

    Read 2026-09-13 · official site · Report a correction

HIPAABAA-eligible endpoints can process PHI once a Business Associate and Healthcare Addendum is executed; live Web Search is not HIPAA eligible.
Sources (1) →
  • Data controls in the OpenAI platform - OpenAI API docs ↗

    “For customers who have executed an OpenAI Business Associate and Healthcare Addendum, once your org ID is provisioned with Eyes Off, BAA-eligible endpoints can be used for processing PHI [...] Web Search with live internet access is not HIPAA eligible and is not covered by a BAA.”

    Read 2026-09-13 · official docs · Report a correction

GDPR / data residencyPer-project data residency stores customer content at rest in a chosen region; non-US regions need abuse-monitoring approval and a Modified Retention amendment.
Sources (1) →
  • Data controls in the OpenAI platform - OpenAI API docs ↗

    “Data residency controls are a project configuration option that allow you to configure the location of infrastructure OpenAI uses to provide services. [...] To use data residency with any region other than the United States, you must be approved for abuse monitoring controls, and execute a Modified Retention amendment.”

    Read 2026-09-13 · official docs · Report a correction

PCI DSSPCI DSS v4.0.1: compliance maintained for the components of ChatGPT that support delegated payment processing. Stated for ChatGPT delegated payment processing; the API platform is not named in the PCI scope.
Sources (1) →
  • OpenAI Trust Portal ↗

    “OpenAI now maintains PCI-DSS compliance for the components of ChatGPT that support delegated payment processing, ensuring secure handling of payment information for supported merchant transactions.”

    Read 2026-09-13 · official site · Report a correction

Encryption at restData protection controls include encryption of your data in transit and at rest, per the applicable ChatGPT Enterprise or API organization settings. Stated in OpenAI's Codex HIPAA configuration guide; no separate API-docs statement.
Sources (1) →
Customer-managed keysEnterprise Key Management encrypts customer content with your own KMS keys; BYOK via AWS KMS, Google Cloud or Azure Key Vault.
Sources (1) →
  • Data controls in the OpenAI platform - OpenAI API docs ↗

    “Enterprise Key Management (EKM) allows you to encrypt your customer content at OpenAI using keys managed by your own external Key Management System (KMS). [...] OpenAI supports Bring Your Own Key (BYOK) encryption with external accounts in AWS KMS, Google Cloud (GCP), and Azure Key Vault.”

    Read 2026-09-13 · official docs · Report a correction

Role-based access controlRole-based access control with preset and custom roles, groups syncable from an identity provider via SCIM, scoped at organization or project level.
Sources (1) →
  • Role-based access control (RBAC) - OpenAI API docs ↗

    “Role-based access control (RBAC) lets you decide who can do what across your organization and projects [...] Create custom roles with the exact permissions you need [...] Groups : Collections of users you can assign roles to. Groups can be synced from your identity provider (via SCIM)”

    Read 2026-09-13 · official docs · Report a correction

Private networking / VPCAzure Private Link to regional OpenAI API endpoints; not self-service, and incompatible with IP allowlist controls or mutual TLS.
Sources (1) →
  • Private Link - OpenAI API docs ↗

    “OpenAI Private Link lets Azure workloads reach regional OpenAI API endpoints through Azure Private Link instead of connecting directly to public API endpoints. [...] Private Link is currently not self-service.”

    Read 2026-09-13 · official docs · Report a correction

Dedicated infrastructureScale Tier offers more predictable capacity on eligible models for enterprise pay-as-you-go traffic; Reserved Tier covers GPT-5.6 and later models.
Sources (1) →
  • Rate limits - OpenAI API docs ↗

    “Enterprise customers whose pay-as-you-go traffic routinely hits ramp-rate limits can consider Scale Tier for more predictable capacity on eligible models. For GPT-5.6 and later models, see Reserved Tier”

    Read 2026-09-13 · official docs · Report a correction

When these numbers change, hear about it. Sources are re-checked monthly; a repricing goes out as a short note.

Confirm by email; unsubscribe from any issue. Your address goes to Kit and nowhere else — what we do with it.

OpenAI in the news

Headlines from other publications, not checked by us · collected

Selected by a search on this company, refreshed weekly. Nobody here has verified these claims — where something material is confirmed, it is written into the page above with its source, the way every other fact on this site is.

Comments

Every page here is sourced and dated.