📍 Independent. Unsponsored. Reliable.

OpenAI Marketplace: How Enterprises Can Spend Their Commitment on Partner Apps

The OpenAI Marketplace lets eligible enterprise customers apply part of their existing OpenAI spending commitment to approved partner software, including Adobe, Figma, Salesforce and ServiceNow. It launched at DevDay on September 29, 2026 with 32 …

Storefront with a coin stack and app tiles, illustrating the OpenAI Marketplace

The OpenAI Marketplace lets eligible enterprise customers apply part of their existing OpenAI spending commitment to approved partner software, including Adobe, Figma, Salesforce and ServiceNow. It launched at DevDay on September 29, 2026 with 32 launch partners. You still sign with and pay the partner. OpenAI checks eligibility and tracks the drawdown against your commitment.

That sounds simple, but the details decide whether it helps you. As of September 30, 2026, OpenAI has not published discount terms, a cap on how much commitment you can redirect, or how usage is metered. This guide separates what is confirmed from what is still unknown, and it shows L&D and HR-tech buyers what to do next.

The marketplace is one of several enterprise announcements from the event. For the wider picture, read our OpenAI DevDay 2026 announcements guide and the OpenAI DevDay 2026 recap.

What is the OpenAI Marketplace?

The OpenAI Marketplace is a program that lets eligible enterprise customers apply part of their existing OpenAI commitment toward approved partner software. It works more like a procurement route than a consumer app store. You buy from the partner, and OpenAI counts the eligible spend against the amount you already agreed to spend with it.

OpenAI describes it in its DevDay 2026 recap as giving customers “more flexibility in how they work with AI.” The OpenAI Marketplace page shows a partner directory, a “Request to buy” form and a waitlist for vendors that want to join.

The rollout looks staged. ElevenLabs said in its launch post that the directory was live with purchasing “available soon.” OpenAI’s help article calls this an “initial rollout managed directly with OpenAI and participating partners.”

Do not confuse it with Sign in with ChatGPT. That feature lets individual users spend their own plan allowance inside third-party apps, and we cover it in our Sign in with ChatGPT guide.

How does OpenAI Marketplace billing and commitment drawdown work?

Billing is split in two. You sign a contract with the partner, and the partner invoices you, at least during the initial rollout. OpenAI confirms that you and the specific product qualify, then reconciles the purchase against your commitment. There is no OpenAI checkout, and no self-service purchasing route has been described.

OpenAI’s help article on the marketplace says purchases “are arranged directly with OpenAI and participating partners.” Its marketplace page adds: “You contract with and pay the partner directly; OpenAI reconciles use of your commitment under the program’s terms.”

In practice, expect a sequence like this:

  1. You find a product in the directory and raise a request with the partner or your OpenAI account team.
  2. OpenAI and the partner confirm whether you and the proposed purchase qualify.
  3. You sign the partner’s contract and receive the partner’s invoice.
  4. OpenAI reconciles the eligible amount against your commitment.

The unclear part is metering. OpenAI has not said whether drawdown happens per invoice, per usage period or on a schedule, or how a partial drawdown appears on your OpenAI statements. Ask for a sample statement before you sign anything.

Who are the OpenAI Marketplace launch partners?

OpenAI says the marketplace launched with 32 partners. Its recap names 12, including Adobe, Figma, Salesforce, ServiceNow, Harvey, Legora, CrowdStrike, Palo Alto Networks and Baseten. The marketplace page lists the full set of 32 names, spanning creative, customer experience, legal, security, developer and data tools.

The table shows the groups named in the recap, then the remaining names from the marketplace page. OpenAI does not categorise that last row.

Group Partners Source
Creative Adobe, Figma OpenAI DevDay recap
Customer experience Sierra, Decagon, HubSpot, Salesforce, ServiceNow OpenAI DevDay recap
Legal Harvey, Legora OpenAI DevDay recap
Cybersecurity CrowdStrike, Palo Alto Networks OpenAI DevDay recap
Open-source models Baseten OpenAI DevDay recap
Other names on the page Basis, CodeRabbit, Datadog, DevRev, ElevenLabs, Factory, Glean, Greptile, Hex, Higgsfield, Hyperagent, Lovable, Manus, Notion, Ramp, Replit, Rogo, Runway, Vercel, Zendesk OpenAI Marketplace page (as of September 30, 2026)

Two cautions. First, some early summaries listed Canva among the launch partners. The OpenAI marketplace page does not show Canva, so we have left it out. Second, a partner name is not a product guarantee. OpenAI states that eligibility “applies to specific products; not every product from a listed partner qualifies.”

Several partners have published their own announcements, including DevRev, Sierra and Baseten.

Who is eligible to use the OpenAI Marketplace?

Only eligible enterprise customers can use it. OpenAI says eligibility depends on your OpenAI agreement, the partner product and the program’s terms. Teams on ChatGPT Plus, Pro or standard Business plans without an enterprise commitment have no stated route in. The practical test is whether you hold a committed spend agreement with OpenAI.

Reports differ on geography and status. The Next Web describes the marketplace as open to “eligible US enterprise customers” with 32 beta partners in the United States. OpenAI’s own pages, as we read them, mention neither a US restriction nor a beta label.

If you operate outside the US, treat availability as unconfirmed until your account team says otherwise. OpenAI also has not disclosed a minimum commitment size. Do not assume a small commitment qualifies, and do not assume it does not.

How does the OpenAI Marketplace compare with AWS, Azure and Google Cloud marketplaces?

The idea is the same: spend a commitment you already made on software you planned to buy anyway. The mechanics differ. On the big cloud marketplaces the cloud provider is generally where you transact. On OpenAI’s, the partner invoices you and OpenAI only reconciles. The cloud rules are also better documented than OpenAI’s.

The table uses public documentation and seller-side guides. Cloud rules change, so confirm each line against your contract.

Feature OpenAI Marketplace AWS Marketplace Microsoft Marketplace (Azure) Google Cloud Marketplace
Commitment used Existing OpenAI commitment AWS spend commitment Microsoft Azure Consumption Commitment (MACC) Google Cloud spend commitment
Who invoices Partner invoices; OpenAI reconciles (initial rollout) Purchase flows through your AWS relationship Microsoft issues the invoice Transaction runs through Marketplace on your billing account
Share that counts “Part” of the commitment; percentage not published Historically capped at about 25% of commitment (2022 to May 2025), per one seller guide; verify current terms 100% of the pretax amount for eligible offers Dollar-for-dollar for eligible purchases, per a seller guide
Product rule Eligibility set per customer, product and purchase SaaS must be entirely hosted on AWS since May 2025, per one guide Azure benefit-eligible offer, bought in the Azure portal, primarily hosted in Azure Marketplace transaction on an account with an active spend commitment
Self-service No; arranged with OpenAI and the partner Yes, through the marketplace Yes, through the Azure portal Yes, through the marketplace
Published terms Limited FAQ; no cap or discount stated Public documentation Public documentation Public documentation

Sources for the cloud columns: Microsoft’s MACC marketplace FAQ, a nOps summary of AWS Marketplace commitment rules, Suger’s AWS EDP guide and Suger’s Google Cloud drawdown guide.

OpenAI is not alone among model makers. Techzine reported that Anthropic’s Claude Marketplace, which launched in March 2026, lets customers direct a portion of their Anthropic commitments to third-party software and takes no commission. OpenAI has not published its position on fees.

There is a second procurement route for OpenAI models. Bedrock Managed Agents lets teams run OpenAI agents inside AWS, which keeps that spend in the AWS relationship. Our OpenAI Agents API and computer use guide covers what it does.

Ask For The Statement First

Before you sign a partner contract, ask OpenAI for a mock reconciliation statement showing how a partner invoice appears against your commitment. If they cannot produce one, treat the drawdown mechanics as unproven.

What are the procurement benefits of using an OpenAI commitment on partner software?

The main benefit is budget flexibility. If you committed to more OpenAI spend than your teams will use, the marketplace gives you a way to put part of it toward tools you would buy anyway. It may also shorten onboarding because OpenAI and the partner check eligibility together. Neither benefit is guaranteed until you see the terms.

Unused commitment can be lost or become a painful renewal conversation, and new software may need fresh budget approval. Redirecting committed money can ease both. One outlet, Runtime Wire, put the vendor side plainly: it “creates a sales channel tied to budgets customers have already committed to OpenAI.”

Finance may also get a cleaner view of total AI spend. Individual plan costs are separate, and our ChatGPT pricing guide explains those tiers.

OpenAI calls the partners “approved,” but it has not said what approval covers. Keep your own security, privacy and data-processing review. A marketplace listing is not a substitute for it.

What are the lock-in risks of the OpenAI Marketplace?

The risk is that a commitment meant to add flexibility starts steering your software choices. Partners are built on or work with OpenAI models, so redirected spend can pull you toward OpenAI-centred stacks and longer terms. Discounts may also depend on commitment size. Watch for renewal traps, and for partner contracts that outlast your OpenAI commitment.

Here are the risks worth pricing in:

  • Commitment steering: a tool wins because it is drawable, not because it is best for the job.
  • Term mismatch: a three-year partner contract sitting against a one-year OpenAI commitment.
  • Shrinking model budget: every dollar redirected to a partner is a dollar not spent on OpenAI usage, which can affect your renewal position.
  • Product-level eligibility: a module you rely on may not qualify, or may lose eligibility later.
  • Opaque economics: with no cap or discount disclosed, you cannot yet compare the deal with buying direct.

Analyst commentary on the launch, such as this FourWeekMBA analysis, notes that OpenAI publishes no dollar figure, commitment size or cap. That is a reasonable reading of the public pages, and it is why the questions below matter.

What is still unknown about the OpenAI Marketplace?

As of September 30, 2026, several terms that matter to buyers are unpublished or unclear. These include discount terms, the maximum share of a commitment you can redirect, how usage is metered and reconciled, whether partners pay OpenAI a fee, and whether every country can take part. Treat each as an open question.

One Brazilian outlet, Antihype, reported that OpenAI declined to publish a percentage or commission structure. We could not confirm that from OpenAI, so treat it as reported.

What should you ask your OpenAI account rep about the marketplace?

Ask questions that turn “part of your commitment” into numbers and dates. The most useful are what share you can redirect, which exact partner products qualify, how drawdown is metered, and what happens to unused commitment at renewal. Get the answers in writing, ideally in your order form rather than in an email thread.

Bring this list to the call:

  1. What percentage or dollar amount of our commitment can we apply to partner products?
  2. Which specific SKUs, modules and add-ons from the partner qualify?
  3. How is drawdown metered: per invoice, per period, or on usage?
  4. Does partner spend count fully, or at a reduced rate, against our commitment?
  5. Do we get a discount, or is the partner’s price the same as buying direct?
  6. What happens if the partner product loses eligibility during our term?
  7. Does redirected spend change our renewal tier or discount band?
  8. Is our country and legal entity covered, and is the program still in beta?

Align Contract Terms

Line up the partner’s contract start and end dates with your OpenAI commitment period before signing. If the terms cannot match, negotiate a termination or non-renewal right on the partner side so a shorter OpenAI deal does not strand you.

What does the OpenAI Marketplace mean for L&D and HR-tech software buyers?

For most L&D teams, little changes today. None of the 32 names is a dedicated LMS, TMS or LXP vendor, and no source shows an authoring tool qualifying. Adobe makes authoring software, but OpenAI has not said which Adobe products are eligible. The value is indirect and depends on whether your company holds an OpenAI commitment.

Where it could matter is adjacent tooling. Glean and Notion touch knowledge access and documentation. ElevenLabs sells voice agents that a training team might test for practice conversations. Runway and Adobe cover video and creative work. Whether any qualifying product fits your learning objectives is a separate question from budget source.

Do not let the funding route pick your tools. Start from the learning problem, the SME time available and the Kirkpatrick level you need to measure, then check whether a shortlisted product happens to qualify. Our AI tools guide and the GPT-6 Astra for learning and development piece help with that shortlist.

What if your LMS or authoring vendor is not an OpenAI Marketplace partner?

Do nothing drastic. A vendor missing from the list does not have weaker AI features. It only means its spend cannot be reconciled against an OpenAI commitment today. Keep your current contract, ask the vendor whether it plans to apply to OpenAI’s partner waitlist, and compare total cost rather than budget source.

Ask which model powers the vendor’s AI features and whether usage is bundled or metered. If you are still choosing a platform, our guide to the best AI LMS platforms compares options on capability, and our training software for remote teams guide covers distributed delivery.

What if your LMS or authoring vendor joins the marketplace later?

Confirm before you celebrate. Partner status does not make every product eligible, so ask for the exact modules and AI add-ons that qualify. Then check that your LMS or authoring contract still protects SCORM and xAPI export rights, data ownership and exit terms before you tie the purchase to an OpenAI commitment.

As of September 30, 2026, we found no LMS, TMS or LXP vendor among the launch partners, and we make no claim that any will join. If a vendor tells you it is a partner, ask for the OpenAI listing and written confirmation of which products count.

Conclusion

The OpenAI Marketplace is worth a conversation if your company already holds an OpenAI commitment, especially one you may not fully use. It is not yet worth planning a purchase around, because the discount, cap and metering terms are unpublished.

Your next step this week: find out who owns the OpenAI agreement in your organisation, and send them the eight questions above. Ask for the answers in writing, along with a sample reconciliation statement.

If you are in L&D, add one line to your next vendor review: ask each shortlisted vendor whether it is, or plans to be, an OpenAI Marketplace partner. Then judge the tool on learning outcomes first and funding route second.

FAQ

Q1. What is the OpenAI Marketplace?

The OpenAI Marketplace is a program, launched September 29, 2026, that lets eligible enterprise customers apply part of their existing OpenAI commitment toward approved partner software. You contract with and pay the partner directly, and OpenAI reconciles the eligible spend against your commitment. It is not a self-service app store, and access is arranged through OpenAI and the partner.

Q2. Who are the OpenAI Marketplace launch partners?

OpenAI says 32 partners launched. Its recap names Adobe, Figma, Sierra, Decagon, HubSpot, Salesforce, ServiceNow, Harvey, Legora, Palo Alto Networks, CrowdStrike and Baseten. The marketplace page also lists names such as Notion, Vercel, Zendesk, Datadog, Glean, ElevenLabs and Replit. Not every product from a partner qualifies, so check eligibility before buying.

Q3. Who can use the OpenAI Marketplace?

Only eligible enterprise customers can use it. OpenAI says eligibility depends on your OpenAI agreement, the partner product and the program’s terms. It has not published a minimum commitment size. One outlet describes the launch as limited to US enterprise customers, but OpenAI’s own pages do not say so, so confirm with your account team.

Q4. How does the OpenAI Marketplace differ from AWS or Azure marketplaces?

The goal is similar: use a commitment you already hold to buy software. On AWS and Azure the cloud provider’s marketplace is generally where you transact, and the drawdown rules are publicly documented. On OpenAI’s, the partner invoices you and OpenAI reconciles, and OpenAI has not published a cap, discount or metering method.

Q5. Does the OpenAI Marketplace give you a discount?

OpenAI has not stated any discount. Its published FAQ says only that you may apply part of your commitment and that you pay the partner directly. Any price benefit would come from the partner or from your negotiated terms. Ask your account team and the partner whether marketplace pricing differs from their standard rates.

Q6. Is my LMS vendor an OpenAI Marketplace partner?

As of September 30, 2026, none of the 32 listed partners is a dedicated LMS, TMS or LXP vendor. Adobe is listed, but OpenAI has not said which Adobe products qualify. Check the live marketplace page, and if a vendor claims partner status, ask for written confirmation of the exact products that count against your commitment.

Marcus Reyes

Written by Marcus Reyes

Marcus spent eight years as an LMS integration engineer before moving into technical writing, building SSO configurations, SCORM/xAPI pipelines, and HRIS integrations for mid-size and enterprise deployments. He writes for the people who actually implement these systems, admins, developers, and IT directors, and has little patience for vendor marketing that skips the technical fine print. When he’s not documenting API specs, he’s usually breaking a staging environment on purpose to see what happens.

Table of contents