ChatGPT vs. Meta AI: data protection per plan

not publicly documented (the Enterprise privacy page mentions a DPA only for ChatGPT Business, Enterprise and API; Edu via Student Data Privacy Agreement)

not publicly documented (the Enterprise privacy page mentions a DPA only for ChatGPT Business, Enterprise and API; the Help Center article on Go refers on data protection only to the Privacy Policy, Terms of Use and Data Controls FAQ)

not publicly documented (the Enterprise privacy page mentions a DPA only for ChatGPT Business, Enterprise and API; the Help Center article on Plus refers on data protection only to the Privacy Policy, Terms of Use and Data Controls FAQ)

not publicly documented (the Enterprise privacy page mentions a DPA only for ChatGPT Business, Enterprise and API; Edu via Student Data Privacy Agreement)

yes: Enterprise privacy page: "we are able to execute a Data Processing Addendum (DPA) with customers for their use of ChatGPT Business, ChatGPT Enterprise, and the API"; concluded via the DPA form. OpenAI DPA (Effective 2026-01-01) published.

not publicly documented whether a DPA under Art. 28 GDPR can be concluded: Enterprise privacy page: "For ChatGPT Edu and for Teachers, we process your data pursuant to our Student Data Privacy Agreement." It covers ChatGPT For Teachers via the Education Terms and names only US law as "Privacy Laws" (FERPA, PPRA, supplemental state laws); the GDPR and Art. 28 are not mentioned. The offering is aimed only at US schools and school districts.

yes: Enterprise privacy page: "we are able to execute a Data Processing Addendum (DPA) with customers for their use of ChatGPT Business, ChatGPT Enterprise, and the API"; concluded via the DPA form. OpenAI DPA (Effective 2026-01-01) published.

not publicly documented whether a DPA under Art. 28 GDPR can be concluded: Enterprise privacy page: "For ChatGPT Edu and for Teachers, we process your data pursuant to our Student Data Privacy Agreement." The Student Data Privacy Agreement is a contract between OpenAI, LLC and the customer; in it, OpenAI is considered a FERPA "School Official", and "Privacy Laws" covers only US federal and state law (FERPA, PPRA, supplemental state laws); the GDPR and Art. 28 are not mentioned. The Enterprise privacy page mentions the OpenAI DPA only for Business, Enterprise and API.

yes: Enterprise privacy page: "we are able to execute a Data Processing Addendum (DPA) with customers for their use of ChatGPT Business, ChatGPT Enterprise, and the API"; concluded via the DPA form. OpenAI DPA (Effective 2026-01-01) published.

not publicly documented (the Meta AI Terms of Service (EU) are a consumer contract with Meta Platforms Ireland Limited and refer to Meta's privacy policy for personal data; processor terms are not mentioned there)

yes, if Contributor mode is switched off: section 5 of the terms of use applies to Standard Services 'including where you have changed your Account setting under a Coding Harness Subscription' (then MGPT, Meta as processor). In the preset Contributor mode, the Contributor terms apply (section 6, no personal data).

yes: For Standard Services, the terms of use (section 5.1) incorporate the 'Meta Global Processor Terms' (MGPT, as of 2026-03-20) and the data security terms; Meta acts as processor. For European data, Europe-specific MGPT terms and an addendum on global data transfer apply.

no (the processor terms are incorporated in section 5 only for Standard Services; terms of use 6.2: no sensitive, confidential or personal information may be submitted to discounted services)

Help Center (section Services for individuals): "When you use our services for individuals (Free, Plus, Pro), we may use your content to train our models. You can choose whether your conversations help improve our models. To opt out, turn off Improve the model for everyone under Settings > Data controls in ChatGPT, or select Do not train on my content in our Privacy Portal." Temporary chats: "are not used to improve OpenAI models while they remain temporary". Feedback exception: "If you choose to provide feedback, the entire conversation associated with that feedback may be used to train our models."

Help Center (section Services for individuals): "When you use our services for individuals (Free, Plus, Pro), we may use your content to train our models. You can choose whether your conversations help improve our models. To opt out, turn off Improve the model for everyone under Settings > Data controls in ChatGPT, or select Do not train on my content in our Privacy Portal." Temporary chats: "are not used to improve OpenAI models while they remain temporary". Feedback exception: "If you choose to provide feedback, the entire conversation associated with that feedback may be used to train our models." Help Center article on Go: "OpenAI may use conversations to improve model performance and safety. You can opt out of having your data used for training at any time."

Help Center (section Services for individuals): "When you use our services for individuals (Free, Plus, Pro), we may use your content to train our models. You can choose whether your conversations help improve our models. To opt out, turn off Improve the model for everyone under Settings > Data controls in ChatGPT, or select Do not train on my content in our Privacy Portal." Temporary chats: "are not used to improve OpenAI models while they remain temporary". Feedback exception: "If you choose to provide feedback, the entire conversation associated with that feedback may be used to train our models."

Help Center (section Services for individuals): "When you use our services for individuals (Free, Plus, Pro), we may use your content to train our models. You can choose whether your conversations help improve our models. To opt out, turn off Improve the model for everyone under Settings > Data controls in ChatGPT, or select Do not train on my content in our Privacy Portal." Temporary chats: "are not used to improve OpenAI models while they remain temporary". Feedback exception: "If you choose to provide feedback, the entire conversation associated with that feedback may be used to train our models."

Help Center: "By default, we don't use inputs or outputs from ChatGPT Business, ChatGPT Enterprise, ChatGPT Edu, or our API to improve our models." Enterprise privacy page: "We do not train our models on your data by default". Help Center (Business FAQ): "ChatGPT Business is governed by OpenAI Business Terms and the Enterprise Privacy commitments, and OpenAI does not train on your workspace data."

Help Center: "As with other Business customers, information shared with ChatGPT for Teachers is not used to train our models by default." In addition: "we may use automated systems to classify content for safety and quality."

Help Center: "By default, we don't use inputs or outputs from ChatGPT Business, ChatGPT Enterprise, ChatGPT Edu, or our API to improve our models." Enterprise privacy page: "We do not train our models on your data by default".

Help Center: "By default, we don't use inputs or outputs from ChatGPT Business, ChatGPT Enterprise, ChatGPT Edu, or our API to improve our models." Enterprise privacy page: "We do not train our models on your data by default".

Help Center: "By default, we don't use inputs or outputs from ChatGPT Business, ChatGPT Enterprise, ChatGPT Edu, or our API to improve our models." Enterprise privacy page: "We do not train our models on your data by default". API documentation: "data sent to the OpenAI API is not used to train or improve OpenAI models (unless you explicitly opt in to share data with us)". API organisation owners can enable data sharing (Help Center); only then can members share feedback and the current Playground chat.

Privacy Center 'How Meta uses information for generative AI models' (translated) (q9): training sources include 'interactions with features that are part of AI at Meta' (translated), public information in Meta products, internet content and licensed data. Private messages to friends and family are not used unless they are knowingly shared with the AIs. There is a right to object to the use of interactions with AI at Meta and of public information (via the account or regional forms according to the Privacy Center; applies to future training). In the UK and the EU, no public information from accounts of under-18s.

Subscription docs: Muse Code subscriptions are in Contributor mode by default ('default to Contributor mode'); content is then used to train future Meta models. The mode can be switched off in the settings at any time; thereafter, future content is not used for training.

Terms of use (as of 2026-10-09, 5.1): Meta does not use content from Standard Services to train Meta models. Pricing page: 'In Standard mode, you pay standard prices and your content is not used to train Meta models.' Data commitments: content from paid services is not used for training; customer data (account, roles, profile fields) in no case.

Pricing page: Contributor mode 'at heavily discounted token prices, in exchange for permission to use your content to train future Meta models'. Terms of use (section 6): content of discounted services may be used by Meta for training.

Privacy policy (Europe, 2026-08-24), section 10: "OpenAI processes your Personal Data on servers located outside of the EEA, Switzerland and the UK ... This includes processing and storing your Personal Data in our facilities and servers in the United States or in countries or territories where our affiliates and partners or our vendors and service providers are located." Pricing page: row "Data residency in US, EU, UK, JP, CA, KR, SG, IN, AU, UAE" = No for Free, Go, Plus, Pro.

Privacy policy (Europe, 2026-08-24), section 10: "OpenAI processes your Personal Data on servers located outside of the EEA, Switzerland and the UK ... This includes processing and storing your Personal Data in our facilities and servers in the United States or in countries or territories where our affiliates and partners or our vendors and service providers are located." Pricing page: row "Data residency in US, EU, UK, JP, CA, KR, SG, IN, AU, UAE" = No for Free, Go, Plus, Pro.

Privacy policy (Europe, 2026-08-24), section 10: "OpenAI processes your Personal Data on servers located outside of the EEA, Switzerland and the UK ... This includes processing and storing your Personal Data in our facilities and servers in the United States or in countries or territories where our affiliates and partners or our vendors and service providers are located." Pricing page: row "Data residency in US, EU, UK, JP, CA, KR, SG, IN, AU, UAE" = No for Free, Go, Plus, Pro.

Privacy policy (Europe, 2026-08-24), section 10: "OpenAI processes your Personal Data on servers located outside of the EEA, Switzerland and the UK ... This includes processing and storing your Personal Data in our facilities and servers in the United States or in countries or territories where our affiliates and partners or our vendors and service providers are located." Pricing page: row "Data residency in US, EU, UK, JP, CA, KR, SG, IN, AU, UAE" = No for Free, Go, Plus, Pro.

not publicly documented for Business: the residency page in the Help Center (as of 2026-10-18) names as eligible "Eligible API customers and new ChatGPT Enterprise/Edu customers" (q5), the business data page "Eligible ChatGPT Enterprise, ChatGPT Edu, ChatGPT for Healthcare, and API platform customers" (q11); ChatGPT Business is not named in either. The pricing page lists data residency in ten regions only for the Enterprise plan. The sub-processor list (as of 2026-10-14) names processing locations for Business in the USA and other countries (including EU Member States), without a choice.

not publicly documented: the Help Center refers to sales for data residency and EKM ("You require advanced enterprise features (for example, data residency or EKM)"); the residency page does not name Teachers.

Help Center Data residency (as of 2026-10-18): "Eligible API customers and new ChatGPT Enterprise/Edu customers can choose to have in-scope customer content stored at rest in supported countries." Regions: Australia, Canada, Europe (EEA + Switzerland), India, Japan, Singapore, South Korea, UAE, UK, USA. Inference residency (GPU execution in the region) only for Europe (EEA + Switzerland), USA and UAE, and only with data residency activated in the same region. In scope: Memory, Code Interpreter/Data Analysis artefacts, conversations (text, image, voice), Custom GPTs, files, image generation inputs/outputs, incl. backups/replicas; new features only if listed. Outside the region, data from external integrations (Apps & MCP, Web Search), transient processing steps, workspace metadata and name, billing information and logins, among others, may be stored ("Certain categories of data may be stored outside the chosen data region"); CPU processing such as authentication, routing and analytics may run outside the region. Applies to newly provisioned workspaces; eligibility via sales; at no additional cost for Enterprise and Education.

Help Center Data residency (as of 2026-10-18): "new ChatGPT Enterprise/Edu customers can choose to have in-scope customer content stored at rest in supported countries"; Europe (EEA + Switzerland) is a supported region; inference residency for Enterprise/Edu with data residency activated (Europe, USA, UAE). Exceptions as for Enterprise (including external integrations, workspace metadata, billing information, logins). Eligibility via sales; at no additional cost for Education.

Help Center Data residency: "Eligible API customers ... can choose to have in-scope customer content stored at rest in supported countries". API documentation (Data controls, as of 2026-10-18): data residency is a project setting; "Contact our sales team to see if you're eligible for using data residency controls." Regions with their own endpoints: USA, Europe (EEA + Switzerland, eu.api.openai.com), Australia, Canada, Japan, India, Singapore, South Korea, UK, UAE; inference in the region only for the USA, Europe and the UAE, the other regions storage only; all regions except the USA require Modified Abuse Monitoring or Zero Data Retention. System data (account data, metadata, usage data without customer content) may be processed and stored outside the region. For models released from 2026-03-05, a 10% uplift applies on residency endpoints.

not publicly documented on the retrieved pages.

Region: not publicly documented. According to the docs, subscription features and availability may vary by region.

Region and EU option for the API: not publicly documented. Terms of use (section 6): 'Meta does not receive, determine, or route traffic based on location of your End Users'. The sub-processor list linked in section 5 names data centre operators predominantly in the USA as well as in Denmark, Sweden and Ireland; which of them are used for the Model API is not stated. Model docs: 'Model availability can vary by country' (reference to the Geographic Use Policy).

Region and EU option: not publicly documented.

• not publicly documented for consumer plans (the sub-processor list, as of 2026-10-14, names API, ChatGPT Enterprise, Edu and Business as its scope; for Plus and Pro it states only for ChatGPT Sites that the listed sub-processors may provide web hosting, infrastructure, moderation and support; reference: https://openai.com/policies/sub-processor-list/)

• not publicly documented for consumer plans (the sub-processor list, as of 2026-10-14, names API, ChatGPT Enterprise, Edu and Business as its scope; for Plus and Pro it states only for ChatGPT Sites that the listed sub-processors may provide web hosting, infrastructure, moderation and support; reference: https://openai.com/policies/sub-processor-list/)

• not publicly documented for consumer plans (the sub-processor list, as of 2026-10-14, names API, ChatGPT Enterprise, Edu and Business as its scope; for Plus and Pro it states only for ChatGPT Sites that the listed sub-processors may provide web hosting, infrastructure, moderation and support; reference: https://openai.com/policies/sub-processor-list/)

• not publicly documented for consumer plans (the sub-processor list, as of 2026-10-14, names API, ChatGPT Enterprise, Edu and Business as its scope; for Plus and Pro it states only for ChatGPT Sites that the listed sub-processors may provide web hosting, infrastructure, moderation and support; reference: https://openai.com/policies/sub-processor-list/)

• According to the list (as of 2026-10-14, applies to API, ChatGPT Enterprise, Edu, Business): OpenAI companies OpenAI, OpCo, LLC and OAI International, Inc. (USA), OpenAI Ireland Ltd. (Ireland), OpenAI UK Ltd., OpenAI Japan Ltd. (standard contractual clauses between group companies)
• Cloudflare, Ltd. (CDN; data centre closest to the end user; for requests to us.api.openai.com or eu.api.openai.com, TLS termination in the selected region via Cloudflare Regional Services; also web hosting for ChatGPT Sites in the USA for Enterprise, Edu, Business)
• Microsoft Corporation (cloud infrastructure; including Belgium, Denmark, Germany, France, Ireland, Italy, the Netherlands, Norway, Poland, Sweden, Switzerland, Spain, the UK, the USA and others)
• CoreWeave, Inc. (cloud infrastructure; Norway, Spain, Sweden, UK, USA)
• Oracle Cloud Infrastructure (cloud infrastructure; Brazil, Japan, Malaysia, the Netherlands, the UK, the USA)
• Google Cloud Platform (cloud infrastructure; Finland, Japan, the Netherlands, Norway, Singapore, the UK, the USA)
• Amazon Web Services, Inc. (cloud infrastructure; USA)
• Cerebras (cloud infrastructure; USA, Canada)
• Snowflake, Inc. (data warehousing; USA; for the API except with Zero Data Retention)
• TaskUs, LLC (customer support; for API and Business moderation of content; for Enterprise, Edu and Business moderation of GPTs; Philippines)
• Salesforce (customer support; including Germany, France, Ireland, Italy, the Netherlands, Sweden, Switzerland, the UK, the USA and others)
• Pylon Labs (customer support; USA; API, Enterprise, Edu only; at the customer's choice, premium support only)
• Accenture International Limited (customer support; for API/Business also moderation of content; USA, Canada, Philippines)
• Fivetran, Inc. (ETL; USA)
• Cinder Technologies, Inc. (platform for content or GPT moderation; USA; for API except where Zero Data Retention is used)
• WorkOS, Inc. (cross-domain identity management; USA; at the customer's choice; Enterprise, Edu, Business)
• Merge API, Inc. (infrastructure management, only with activated Connectors; including France, Ireland, Sweden, the UK, the USA and others; at the customer's choice)

• not publicly documented (the sub-processor list, as of 2026-10-14, does not name ChatGPT for Teachers; according to the Help Center, authorised employees and "specialized third-party contractors" have access to stored conversations to review for abuse)

• According to the list (as of 2026-10-14, applies to API, ChatGPT Enterprise, Edu, Business): OpenAI companies OpenAI, OpCo, LLC and OAI International, Inc. (USA), OpenAI Ireland Ltd. (Ireland), OpenAI UK Ltd., OpenAI Japan Ltd. (standard contractual clauses between group companies)
• Cloudflare, Ltd. (CDN; data centre closest to the end user; for requests to us.api.openai.com or eu.api.openai.com, TLS termination in the selected region via Cloudflare Regional Services; also web hosting for ChatGPT Sites in the USA for Enterprise, Edu, Business)
• Microsoft Corporation (cloud infrastructure; including Belgium, Denmark, Germany, France, Ireland, Italy, the Netherlands, Norway, Poland, Sweden, Switzerland, Spain, the UK, the USA and others)
• CoreWeave, Inc. (cloud infrastructure; Norway, Spain, Sweden, UK, USA)
• Oracle Cloud Infrastructure (cloud infrastructure; Brazil, Japan, Malaysia, the Netherlands, the UK, the USA)
• Google Cloud Platform (cloud infrastructure; Finland, Japan, the Netherlands, Norway, Singapore, the UK, the USA)
• Amazon Web Services, Inc. (cloud infrastructure; USA)
• Cerebras (cloud infrastructure; USA, Canada)
• Snowflake, Inc. (data warehousing; USA; for the API except with Zero Data Retention)
• TaskUs, LLC (customer support; for API and Business moderation of content; for Enterprise, Edu and Business moderation of GPTs; Philippines)
• Salesforce (customer support; including Germany, France, Ireland, Italy, the Netherlands, Sweden, Switzerland, the UK, the USA and others)
• Pylon Labs (customer support; USA; API, Enterprise, Edu only; at the customer's choice, premium support only)
• Accenture International Limited (customer support; for API/Business also moderation of content; USA, Canada, Philippines)
• Fivetran, Inc. (ETL; USA)
• Cinder Technologies, Inc. (platform for content or GPT moderation; USA; for API except where Zero Data Retention is used)
• WorkOS, Inc. (cross-domain identity management; USA; at the customer's choice; Enterprise, Edu, Business)
• Merge API, Inc. (infrastructure management, only with activated Connectors; including France, Ireland, Sweden, the UK, the USA and others; at the customer's choice)

• According to the list (as of 2026-10-14, applies to API, ChatGPT Enterprise, Edu, Business): OpenAI companies OpenAI, OpCo, LLC and OAI International, Inc. (USA), OpenAI Ireland Ltd. (Ireland), OpenAI UK Ltd., OpenAI Japan Ltd. (standard contractual clauses between group companies)
• Cloudflare, Ltd. (CDN; data centre closest to the end user; for requests to us.api.openai.com or eu.api.openai.com, TLS termination in the selected region via Cloudflare Regional Services; also web hosting for ChatGPT Sites in the USA for Enterprise, Edu, Business)
• Microsoft Corporation (cloud infrastructure; including Belgium, Denmark, Germany, France, Ireland, Italy, the Netherlands, Norway, Poland, Sweden, Switzerland, Spain, the UK, the USA and others)
• CoreWeave, Inc. (cloud infrastructure; Norway, Spain, Sweden, UK, USA)
• Oracle Cloud Infrastructure (cloud infrastructure; Brazil, Japan, Malaysia, the Netherlands, the UK, the USA)
• Google Cloud Platform (cloud infrastructure; Finland, Japan, the Netherlands, Norway, Singapore, the UK, the USA)
• Amazon Web Services, Inc. (cloud infrastructure; USA)
• Cerebras (cloud infrastructure; USA, Canada)
• Snowflake, Inc. (data warehousing; USA; for the API except with Zero Data Retention)
• TaskUs, LLC (customer support; for API and Business moderation of content; for Enterprise, Edu and Business moderation of GPTs; Philippines)
• Salesforce (customer support; including Germany, France, Ireland, Italy, the Netherlands, Sweden, Switzerland, the UK, the USA and others)
• Pylon Labs (customer support; USA; API, Enterprise, Edu only; at the customer's choice, premium support only)
• Accenture International Limited (customer support; for API/Business also moderation of content; USA, Canada, Philippines)
• Fivetran, Inc. (ETL; USA)
• Cinder Technologies, Inc. (platform for content or GPT moderation; USA; for API except where Zero Data Retention is used)
• WorkOS, Inc. (cross-domain identity management; USA; at the customer's choice; Enterprise, Edu, Business)
• Merge API, Inc. (infrastructure management, only with activated Connectors; including France, Ireland, Sweden, the UK, the USA and others; at the customer's choice)

• According to the list (as of 2026-10-14, applies to API, ChatGPT Enterprise, Edu, Business): OpenAI companies OpenAI, OpCo, LLC and OAI International, Inc. (USA), OpenAI Ireland Ltd. (Ireland), OpenAI UK Ltd., OpenAI Japan Ltd. (standard contractual clauses between group companies)
• Cloudflare, Ltd. (CDN; data centre closest to the end user; for requests to us.api.openai.com or eu.api.openai.com, TLS termination in the selected region via Cloudflare Regional Services; also web hosting for ChatGPT Sites in the USA for Enterprise, Edu, Business)
• Microsoft Corporation (cloud infrastructure; including Belgium, Denmark, Germany, France, Ireland, Italy, the Netherlands, Norway, Poland, Sweden, Switzerland, Spain, the UK, the USA and others)
• CoreWeave, Inc. (cloud infrastructure; Norway, Spain, Sweden, UK, USA)
• Oracle Cloud Infrastructure (cloud infrastructure; Brazil, Japan, Malaysia, the Netherlands, the UK, the USA)
• Google Cloud Platform (cloud infrastructure; Finland, Japan, the Netherlands, Norway, Singapore, the UK, the USA)
• Amazon Web Services, Inc. (cloud infrastructure; USA)
• Cerebras (cloud infrastructure; USA, Canada)
• Snowflake, Inc. (data warehousing; USA; for the API except with Zero Data Retention)
• TaskUs, LLC (customer support; for API and Business moderation of content; for Enterprise, Edu and Business moderation of GPTs; Philippines)
• Salesforce (customer support; including Germany, France, Ireland, Italy, the Netherlands, Sweden, Switzerland, the UK, the USA and others)
• Pylon Labs (customer support; USA; API, Enterprise, Edu only; at the customer's choice, premium support only)
• Accenture International Limited (customer support; for API/Business also moderation of content; USA, Canada, Philippines)
• Fivetran, Inc. (ETL; USA)
• Cinder Technologies, Inc. (platform for content or GPT moderation; USA; for API except where Zero Data Retention is used)
• WorkOS, Inc. (cross-domain identity management; USA; at the customer's choice; Enterprise, Edu, Business)
• Merge API, Inc. (infrastructure management, only with activated Connectors; including France, Ireland, Sweden, the UK, the USA and others; at the customer's choice)

• According to the list linked in the terms of use (title 'Meta Horizon Managed Solutions – List of Sub-processors' (translated), valid from 2023-11-13): data centre infrastructure by Andale, Greater Kudu, Goldframe, Morning Hornet, Offprints, Omanyte, Paile, Raven Northbrook, Scout Development, Siculus, Sidecat, Stadion, Starbelt, Vitesse, Winner (d/b/a Ernst), Woolhawk (each USA), Cassin Networks (Denmark), Pinnacle Sweden (Sweden), Runways Information Services (Ireland)
• Meta subsidiaries according to the same list: Meta Platforms, Inc. and Meta Operations, LLC (USA), Meta Platforms Ireland Limited (Ireland), Meta Canada, Meta France, Meta Germany, Meta Italy, Meta Netherlands, Meta Poland, Meta Spain, Meta UK, Meta Singapore, Meta Platforms International Operations Philippines

€0 per month (pricing page, de-DE rendered in the browser; according to the FAQ, Free is available to everyone) (as of 2026-10-18)

€8 per month (pricing page, de-DE rendered in the browser; no statement whether incl. VAT). Help Center: purchases in USD, local-currency billing only in some countries; monthly billing only, no annual plan; no USD amount is stated there (as of 2026-10-18)

€23 per month (pricing page, de-DE rendered in the browser; no statement whether incl. VAT). Help Center: "$20/month (billed monthly)"; no annual plan (as of 2026-10-18)

Pro 100: USD 100, Pro 200: USD 200, Pro 500: USD 500 per month (Help Center; Ultrafast models only in Pro 500; monthly billing only, no annual plan). Pricing page, de-DE rendered in the browser: from €103 per month; euro amounts for Pro 200 and Pro 500 are not shown there (as of 2026-10-18)

Standard seat €21 per user per month billed annually, €26 billed monthly; Premium seat (5x usage, no five-hour limit) €105 or €130 respectively (pricing page, de-DE rendered in the browser). Help Center: Standard USD 20 annually or USD 25 monthly, Premium USD 100 or USD 125 per user per month; "Prices listed here are in USD and may vary by country and currency". At least 2, at most 200 paid seats (as of 2026-10-18)

free for verified US K–12 teachers until June 2028 (Help Center, as of 2026-10-18); pricing possible thereafter with advance notice. The FAQ on the pricing page (de-DE) states June 2027 instead.

Price on request (pricing page: custom pricing via sales; annual plan; billing by credits or by tokens available) (as of 2026-10-18)

not publicly documented (the pricing page describes ChatGPT Edu as a low-cost plan for universities, without an amount; contact via sales) (as of 2026-10-18)

Price not publicly documented (usage-based API prices not evaluated from a primary source)

not publicly documented on the retrieved pages

Flat monthly price in USD (Help Center 'What is a Muse Code subscription?'): Everyday Usage USD 5/month, High Usage USD 15/month (5x usage), Power Usage USD 50/month (20x), Maximum Usage USD 200/month (80x); automatic monthly renewal; not stated whether VAT is included. According to the docs, benefits and availability may vary by region.

Per 1M tokens, in USD (pricing page): Muse Spark input USD 1.25, cached input USD 0.15, output USD 4.25; Muse Wonder input USD 4.00, cached input USD 0.20, output USD 20.00; no surcharge for long context. Muse Image USD 0.01 per image; Muse Voice Transcribe USD 0.18 per hour (ZDR at the same price); web search grounding USD 2.50 per 1,000 search queries; SAM 3.1 USD 2.50 per 1,000 images, USD 0.20 per 1,000 video frames; no minimum commitment. Taxes are added (terms of use 13.1: the customer bears applicable sales, value added or similar taxes).

Per 1M tokens, in USD (pricing page): input USD 0.10, cached input USD 0.002, output USD 0.20. Taxes are added (terms of use 13.1: the customer bears applicable sales, value added or similar taxes).

yes (according to the pricing page: unlimited text chats with GPT-6, limited uploads, image creation, voice chats)

yes (Free plus more messages with tools, uploads, image creation, voice chats, longer Memory; according to the pricing page "This plan may include ads")

yes (Go plus advanced GPT-6 reasoning models, more messages/uploads, Deep Research, Projects, Codex, ChatGPT Work)

yes (Plus plus three usage tiers, Pro reasoning with GPT-6 Astra, Codex/Work sessions, Dot, maximum Deep Research)

yes (Business workspace with workspace admin control of retention according to the Enterprise privacy page; feature scope not evaluated in the retrieved pricing text)

yes (according to the Help Center, a self-serve workspace for verified US K–12 teachers and school staff, not for students; including file uploads, deep research, voice mode, image generation, memory, Custom GPTs, admin features with SSO and RBAC)

yes (according to Enterprise Privacy: SAML SSO, Compliance API/audit logs, RBAC, data residency)

yes (according to the Enterprise privacy page part of the business products; feature scope not evaluated in the retrieved page text)

API (models via Platform); feature scope not the subject of this entry

yes (Meta AI assistant in Meta products and on meta.ai; part of 'AI at Meta' (translated) according to the Privacy Center)

yes (coding agent Muse Code with Muse Spark, Muse Wonder and Muse Image; Power and Maximum additionally Muse Wonder Full; use only via the Muse Code CLI)

yes (models muse-wonder-1.0 ('in preview' according to the model docs), muse-spark-1.3, -1.2, -1.1; image model Muse Image, Muse Voice Transcribe, SAM 3.1; web search grounding)

yes (muse-spark-1.3-contributor, muse-spark-1.2-contributor; not for Muse Wonder and Muse Voice Transcribe)