File path: /Users/forgebot/.hermes/cache/scratch/openai-plans-2026-09-29.html
ForgeFX / Software Plans / Checked September 29, 2026

OpenAI Plans & Team Access

What each tier buys, what it does not, and what still needs an account-level check.

ForgeFX Simulations
The key distinction: ChatGPT Business (formerly Team) can mix Standard and Premium seats. Premium buys higher usage—not a separate set of core features. Personal Plus/Pro subscriptions and API billing are separate from workspace membership. [3]
Business Standard
$20 / $25

Per seat/month: annual commitment / monthly billing.

Business Premium
$100 / $125

Per seat/month: annual commitment / monthly billing. 5× Standard usage; no 5-hour limit.

ForgeFX Account Audit
Access blocked

ChatGPT and its admin route stopped at a browser check. No roster, seat assignments, or invoice verified.

Plan & Feature Comparison

USD list prices, before applicable taxes. Annual figures are monthly equivalents billed annually. Regional prices, rollout, admin policy, and usage limits can change access. Sources [1–6].

Feature / planFreeGoPlusProBusiness StandardBusiness PremiumEnterprise
Price per person$0$8/mo US*$20/mo$100/mo;
$200 tier†
$20 annual /
$25 monthly
$100 annual /
$125 monthly
Custom quote
Best fitOccasional useMore everyday toolsIndividual advanced workHeavy individual useManaged team baselineHeavy users within a managed teamEnterprise controls and contracts
Advanced reasoningLimited; no Astra/Sol listedThink via Luna; no Sol/AstraAstra + Sol listedExpanded Astra; Pro reasoningFrontier models; flexible limits‡Same access; higher limits‡Flexible model access‡
ChatGPT WorkLimited, desktopLimited, desktopDesktop, web, mobileDesktop, web, mobileIncludedIncludedIncluded
CodexLimitedLimitedIncluded; expanded over FreeExpanded / higher usageIncludedIncluded; higher usageIncluded; contract limits
Files, images, data analysisLimitedMore than FreeExpandedHighest personal access; safeguards applyIncluded; rate limitsSame tools; higher usageIncluded; flexible limits
Deep researchLimitedLimitedExpandedMaximum personal accessIncludedIncludedIncluded
Projects / custom GPT creation / tasksProjects only of these threeAll threeAll threeAll threeAll threeAll threeAll three
Office / Sheets extensionsLimitedLimitedIncludedExpandedExcel, PowerPoint, Sheets listed§Same as Standard§Excel, PowerPoint, Sheets listed§
Company knowledge / workspace agentsNoNoNoNoYesYesYes
Team admin / billing / SAML SSONoNoNoNoYesYesYes
SCIM / EKM / granular RBAC / residencyNoNoNoNoNoNoYes; contract/configuration applies
Training on your contentOpt out availableOpt out availableOpt out availableOpt out availableNo, by defaultNo, by defaultNo, by default
AdsMay appearMay appearAd-freeAd-freeAd-freeAd-freeAd-free
API spend included?NoNoNoNoNoNoSeparate commercial terms; do not assume included

* Go’s $8 US price is from OpenAI’s January launch announcement; its current help page redirects price checks to localized checkout. The pricing-page extraction did not expose consumer dollar amounts. Plus and Pro amounts are corroborated by current Plus help [4].

† OpenAI’s help page still says new $200 Pro subscriptions are paused from September 10. A September 29 post by OpenAI’s Thibault Sottiaux announces reopening September 30 with changed usage accounting, roughly half the former API-dollar equivalent, and no 5-hour limit. This is an announced change, not verified checkout availability. Do not budget using the old “20×” allowance. [4,7]

‡ OpenAI’s business comparison still lists GPT-5.6-family rows while its personal page lists Astra. These pages are not fully synchronized. Exact ForgeFX model access needs its live model picker; absence of an Astra row is not proof of exclusion. “Unlimited” always has abuse safeguards and does not make every tool uncapped.

§ The personal comparison explicitly names Word too; the business comparison omits it. Word entitlement for the actual workspace needs verification. The Go help page and pricing table also disagree on voice limits. Business marketing mentions usage analytics while its enterprise-style analytics-dashboard row says “No”: basic usage/spend visibility is not the full Enterprise dashboard.

Who Is On The ForgeFX Team?

Not yet retrievable in this session. Both the ChatGPT homepage and direct admin route returned “Just a moment…” without account content. That establishes an access blocker—not an empty team and not a lack of administrator permissions.

What The Audit Needs

Authenticated ForgeFX workspace access: member list (including pending invites), role, assigned Standard/Premium/legacy seat, purchased capacity, billing cadence, and any credits or spend caps.

What It Cannot Infer

A workspace roster does not establish each person’s separate personal Plus/Pro subscription. An OpenAI API organization and its usage tier are not the ChatGPT team subscription.

How To Finish

Use an authenticated workspace owner/admin session, or provide exports/screenshots of Members and Billing with payment details hidden. Reconcile every member and seat against billed capacity; keep the roster out of this link-accessible report.

Business requires at least two Standard/Premium seats combined. Legacy Codex-only seats can exist only in eligible older workspaces; do not assume ForgeFX has them. Roles such as owner/admin/member describe permissions, not a paid tier. [3]

Rumor & Unconfirmed Watchlist

These rows are deliberately separate from purchasable plans. They are not a recommendation to buy or a promise of access.

ClaimEvidence / statusWhat Is Still UnknownPlanning Treatment
$500/month “Pro Max”REPORTED LEAK
TestingCatalog reports unreleased subscription references; Android Authority corroborates the report. [8,9]
Official launch, final price, eligibility, usage allowances.Not an official purchasable tier in this comparison.
“Fastest Work and Codex,” 100GB storageUNRELEASED UI CLAIM
Reported Pro Max interface copy; 100GB reported by Android Authority. [9]
Whether these benefits ship and how they differ from current Pro.Do not treat as an entitlement or storage commitment.
Cerebras powers Pro Max speedSPECULATION
TestingCatalog connects “Fastest” wording with existing fast-inference infrastructure. [8]
No confirmed direct link between this rumored plan and Cerebras.No guaranteed speed multiplier or hardware promise.

Practical Recommendation

First reconcile the workspace roster and usage. Standard is the sensible managed-team baseline; reserve Premium for people repeatedly hitting limits. Choose Enterprise for governance and contract requirements—not because it sounds like the next speed setting. Keep any existing $200 Pro subscriptions unchanged until the new terms and actual checkout are clear. This is a plan-selection recommendation, not a finding about ForgeFX’s current purchases.

Sources & Evidence Boundaries

[4] OpenAI: Plus pricing and Pro pause notice
[5] OpenAI: Go access and restrictions
[6] OpenAI: Go US price and ad-free paid tiers
[7] OpenAI executive: announced Pro $200 changes
[8] TestingCatalog: Pro Max leak and infrastructure speculation
[9] Android Authority: September 29 Pro reporting