Choosing Claude for work? Jump to chat, delegated tasks, Office add-ins or API before comparing plan capacity.
Which AI subscription includes the model you want? This AI Model Access Map tracks the current consumer-plan and product signals for popular ChatGPT, Claude, Gemini, and Grok models. It separates model visibility from usage limits, feature rollouts, workspace policies, and API billing so you can compare access without turning a marketing label into a guarantee.
How to use this AI model access map
Start by filtering the provider, the product surface, or the access signal. “Included / listed” means the provider currently lists the model or feature at that plan level. “Expanded paid access” means a paid tier adds more model choice, capacity, priority, or quota. “Higher-tier feature” marks an entitlement such as a Pro-only model or an Ultra-only mode. “Separate product” is used when consumer chat and API/Console billing are not the same thing.
- Find the model family or feature you care about.
- Read the plan map, then open the official source on the card.
- Check the live model picker or plan page for your country, age, account, workspace, and platform.
- Compare limits and price separately; access does not mean unlimited usage.
HiseHub Model Access Map
Which AI plan actually includes the model you want?
Compare consumer-plan access for popular ChatGPT, Claude, Gemini, and Grok models and features. The map separates model availability from usage limits, rollouts, and API billing so a plan name does not create a false promise.
Using Grok? Compare three access routes
Standalone SuperGrok
Choose a direct Grok web or mobile subscription if you want the Grok app itself. Check the buying account, live model picker and weekly Usage panel; a paid tier is not unlimited use.
Grok plans and usage ↗X subscription benefit
Already paying for X? Check your tier before buying again. X lists SuperGrok under Premium+; benefits on Grok require the correctly linked X account. Do not assume every X tier or sign-in has the same access.
X tier benefits ↗ · Account linkage ↗xAI API
Building an integration? Check the Console team, API key, model access and API billing. Sharing a sign-in with Grok does not combine the bills or turn a consumer subscription into API credit.
Separate API billing ↗Compare the two consumer purchase routes, troubleshoot Grok usage limits, or estimate API request costs. These answer different questions; none guarantees your account’s entitlement.
Grok purchase routes checked October 8, 2026. Other cards retain their earlier reviews; no account or model output was tested.
33 entries match your filters
Claude delegated tasks (Cowork rollout)
Feature or mode
Plan mapNew unified experience: Pro / Max rollout on web, desktop and mobile
Migrated accounts can delegate work from a conversation without choosing a separate mode. Cloud tasks can continue after leaving; work using local files or apps still needs Claude Desktop open. Existing Chat / Cowork controls mean the account has not migrated yet. Reviewed October 19, 2026.
Claude for Microsoft 365
Feature or mode
Plan mapPro, Max, Team or Enterprise; supported Office apps and deployment required
Excel, PowerPoint, Word and Outlook are add-in surfaces. Cross-app work needs direct Claude sign-in, open files or mail, and the required toggles/admin permission; it is not supported through an LLM gateway. A Claude plan does not supply the Office app. Reviewed October 19, 2026.
Claude Haiku 5.5 (API)
API / developer product
Plan mapClaude API and documented cloud products; verify the billing channel
Released October 7, 2026 for high-volume work. First-party Standard input / output starts at $0.10 / $0.50 per 1M tokens; prompts above 100,000 tokens use higher rates. An API launch does not establish consumer-plan access. Reviewed October 8, 2026.
GPT-6.1 Sol (Work / Codex)
Work and Codex, not ordinary Chat
Plan mapPlus, Pro, Business, Enterprise and Edu; Free / Go excluded at launch
Select GPT-6.1 Sol in supported Work / Codex clients. Enterprise and Edu access is off by default until an admin enables it. Standard and Fast are listed; this is not Astra Ultrafast. Reviewed October 1, 2026.
GPT-6.1 Sol (API)
API / developer product
Plan mapEligible OpenAI API project and key
Use gpt-6.1-sol. Standard short-input rates are $2 input / $10 output and $0.10 cache reads per million tokens. Above 272K input, higher rates apply to the full request. Reviewed October 1, 2026.
Claude Sonnet 5.5 (API)
API / developer product
Plan mapClaude API or supported cloud platform
Use claude-sonnet-5-5. Standard token rates are $2 input / $10 output per million; API High effort is the default. Subscription limits, cache writes and cloud charges are separate checks. Reviewed October 1, 2026.
ChatGPT Dots
Cloud assistant / feature
Plan mapPro $100 / $200 / $500: adults outside EEA, UK and Switzerland; Business Premium and Enterprise: worldwide
Enterprise requires admin enablement. A Dot’s cloud computer can work while personal devices are off; using a connected personal computer requires that computer online with the app open. Reviewed October 1, 2026.
GPT-6 Astra Ultrafast
Work / Codex speed mode
Plan mapPro $500; eligible usage-billed Enterprise and credit-based Edu; US inference required
Enterprise owners must enable it. Other self-serve plans cannot unlock it by buying credits. Included usage is consumed at 8x Standard; credits / Enterprise pay-as-you-go at 6x. This is not an 8x task-completion guarantee. Reviewed October 1, 2026.
Claude Opus 5.5 (Claude plans)
Claude subscription
Plan mapPro, Max, Team and Enterprise
Anthropic lists Opus 5.5 on paid Claude plans. Check the actual model picker, workspace and usage meter; a listed model is not unlimited usage or API credit. Reviewed September 23, 2026.
GPT-6 Sol (Work / Codex)
Work and Codex, not ordinary Chat
Plan mapPlus, Pro, Business, Enterprise and Edu; client/workspace conditions apply
Choose Sol in Work or Codex when available. Its presence in the API catalog does not enable it in ordinary Chat or override workspace restrictions. Reviewed September 23, 2026.
GPT-6 Luna (Work / Codex)
Work and Codex, not ordinary Chat
Plan mapFree and Go: Standard-speed desktop access subject to rollout; eligible paid plans also list Luna
Plan, client, workspace and rollout determine access. Included limits and Codex credits are not the same as separately billed API usage. Reviewed September 23, 2026.
Claude Opus 5.5 (API)
API / developer product
Plan mapClaude API or supported cloud platform
Claude API model ID claude-opus-5-5. Standard token rates are $4 input and $20 output per million; check cache, Fast and cloud billing separately. Reviewed September 23, 2026.
GPT-6 Sol (API)
API / developer product
Plan mapEligible OpenAI API project and key
Use gpt-6-sol. Standard base rates are $2 input and $10 output per million tokens, with higher full-request rates above 272K input tokens. Reviewed September 23, 2026.
GPT-6 Luna (API)
API / developer product
Plan mapEligible OpenAI API project and key
Use gpt-6-luna. Standard base rates are $0.10 input and $0.50 output per million tokens; cache, longer prompts and premium processing alter cost. Reviewed September 23, 2026.
GPT-5.4 Pro
Consumer chat app
Plan mapChatGPT Pro
OpenAI release notes list GPT-5.4 Pro under the Pro plan options. Confirm the current model picker and regional availability before subscribing.
ChatGPT model picker (Instant, Medium, High)
Consumer chat app
Plan mapFree, Plus, Pro — picker differs
The model picker is rolling out across web and mobile. Workspace, plan, and rollout status can change which controls you see.
OpenAI API models
API / developer product
Plan mapAPI account and project billing
A ChatGPT consumer subscription should not be treated as API credit. Check the API pricing and project billing page separately.
GPT-6 Astra
API / developer product
Plan mapSeparately billed OpenAI API project
The current model documentation lists GPT-6 Astra for API use. Verify model access and rate limits in your own project; this API row does not promise access through a ChatGPT subscription. Reviewed October 2, 2026.
Claude Sonnet 5.5 (Claude plans)
Claude subscription
Plan mapFree baseline; Pro, Max, Team, Enterprise expand access
The pricing table lists Sonnet across consumer and business tiers; Sonnet 5.5 launched September 28. Claude Code eligibility is a separate paid-plan / seat check. Medium is the app default, and no tier promises unlimited model use. Reviewed October 1, 2026.
Claude Opus 4.8
Coding and chat surface
Plan mapPaid plans; check model picker and seat policy
Opus-class access can depend on the plan, seat, surface, rollout, and usage window. Claude Code’s /model view is the live source for that account.
Claude API and Console
API / developer product
Plan mapMax and Team offer claimable monthly API credits; Pro is not eligible
Claude plans and Console remain separate products. Eligible subscribers must link the intended Console organization to claim credits; app usage limits do not become an API balance. Check the linked organization before funding it. Reviewed October 8, 2026.
Claude Fable (Claude plans)
Claude subscription
Plan mapPro: usage credits; Max 5x and Max 20x: 50% of weekly limits
The current pricing table lists Fable differently from other Claude models. Free does not list access; Pro uses usage credits; Max lists a model-specific weekly allowance. Verify the account picker and usage meter, not just the plan name. Reviewed September 28, 2026.
Claude Fable 5.1
API / developer product
Plan mapClaude API; Amazon Bedrock, Google Cloud, Microsoft Foundry, Claude Platform on AWS
Anthropic documents Fable 5.1 on the Claude API and partner clouds. Consumer Claude access, seats, quotas, and API billing remain separate checks.
Gemini 3 Flash-Lite
Gemini Apps
Plan mapNo plan, AI Plus, AI Pro, AI Ultra
Google’s current model-access table lists this model across the tiers; usage limits and feature availability still differ by plan and demand.
Gemini 3.8 Flash
API / developer product
Plan mapGemini API / AI Studio; introductory API price through Dec 31, 2026
Google lists Gemini 3.8 Flash as a stable Gemini API model with introductory pricing. Consumer Gemini app access and plan quotas must be checked separately.
Gemini 3 Flash
Gemini Apps
Plan mapNo plan, AI Plus, AI Pro, AI Ultra
The model is listed for all current tiers, while the plan changes the usage multiplier, context window, and access to extra features.
Gemini 3 Pro
Gemini Apps
Plan mapNo plan, AI Plus, AI Pro, AI Ultra
Google lists Gemini 3 Pro across the current tiers, but advanced thinking levels and feature quotas can be plan-specific.
Deep Think mode
Feature or mode
Plan mapGoogle AI Ultra
Google’s current help page marks Deep Think as an AI Ultra-only feature and notes that advanced thinking consumes more usage.
Grok on X
Consumer chat app
Plan mapX Premium: increased Grok limits; Premium+: higher limits
X describes higher Grok usage limits for Premium and still higher limits for Premium+. Feature availability can vary by platform and location.
Grok 4.7
API / developer product
Plan mapxAI API: grok-4.7; token-billed
xAI lists Grok 4.7 for its API from September 21. API billing is separate from consumer Grok plans; confirm the model picker and account or regional availability before building around it.
Grok 4.6
API / developer product
Plan mapxAI API; regional clusters and pricing apply
xAI documents Grok 4.6 as an API model with a 500K context window. Consumer Grok access, X Premium entitlements, and API billing are separate checks.
SuperGrok via X Premium+
Grok web and mobile
Plan mapX Premium+; confirm the linked account and live entitlement
This is an X subscription benefit, not the only SuperGrok purchase route. Match the linked X account to the paid receipt before buying again; standalone purchases are a different route. Reviewed October 8, 2026.
SuperGrok standalone subscription
Grok web, iOS and Android
Plan mapStandalone SuperGrok plans; no X subscription required for this purchase route
Subscribe directly on Grok web or its mobile apps. Paid plans expand the weekly usage allowance; check the live model picker and Usage panel for your account, product and reset schedule. Reviewed October 8, 2026.
The provider currently lists this model or feature for the named plan family. Limits can still apply.
Paid tiers add model choice, capacity, priority, or a larger quota; they do not mean unlimited use.
Consumer chat access and developer API/Console billing are different products unless the source explicitly says otherwise.
Rollout, region, age, account, workspace, and platform conditions can change what appears in your model picker.
How to read this map
Access is not the same thing as a usage guarantee
This is an editorial snapshot of public provider documentation, not a live view of your account. “Included” means the provider currently describes access at that plan or product level; a model can still be unavailable because of a rollout, region, workspace policy, capacity limit, or feature-specific quota. API billing remains a separate product even when a plan offers eligible credits. For Claude, compare Max’s claimable API credits and actual total cost; a plan name alone does not establish your Console balance.
Before you buy, open the source link on the card and check the provider’s current plan page. If you have already paid and the entitlement is missing, use HiseHub’s AI Subscription Troubleshooter instead of purchasing a second plan.
Sources are linked on every card. Model names, plan names, limits, and availability may change. The latest entry review was October 19, 2026, not a simultaneous recheck of every older entry. Verify the linked live source before purchase. HiseHub is independent and is not affiliated with OpenAI, Anthropic, Google, xAI, or X.
New September 22 models: Compare Opus 5.5, GPT-6 Sol and Luna to distinguish paid chat access, Work/Codex and separately billed API usage.
Model access, usage limits, and rollout status are different questions
A model can appear in a plan description while your account still sees a different picker. Providers roll out models gradually, apply regional or age restrictions, use separate workspace policies, and adjust capacity during demand spikes. Some plans expose the same model family but give different message limits, context windows, thinking modes, file allowances, or priority. That is why this map includes a confidence label and a source link instead of pretending every row is permanent.
For Gemini, for example, a current Google help table can list the same model across multiple tiers while the plan changes quotas, context, and feature availability. For ChatGPT, a model picker may expose different controls across Free, Plus, Pro, or a workspace. For Claude, paid plan, seat, chat, and Claude Code access can diverge. For Grok, X Premium tiers can change usage limits and not every feature is available on every platform or location.
Consumer subscription access is not API access
Consumer subscriptions are designed for a provider’s web or mobile experience. API and developer products use separate projects, keys, rate limits, and billing in many cases. A ChatGPT Plus or Pro subscription should not be treated as API credit. Anthropic explicitly separates paid Claude plans from Console access, and the same distinction is useful when comparing any provider. If you need an integration, open the official API or Console pricing page instead of relying on a consumer plan badge.
What to verify before paying
- Use the exact account and sign-in method that will own the subscription.
- Confirm the plan includes the model or mode, not only the provider brand.
- Check whether the model is available in your country, app version, device, or workspace.
- Read the usage window, context, file, and feature limits before choosing a higher tier.
- For API work, confirm separate billing and whether the model is offered on the developer platform.
When the model is missing after purchase
Do not buy the plan a second time immediately. Check the account, workspace, receipt, app-store route, and provider status first. HiseHub’s AI Subscription Troubleshooter walks through those checks for ChatGPT, Claude, Gemini, and Grok. Once the entitlement is visible, use the map again to understand what the plan should expose and which limits still apply.
Related HiseHub tools
Use AI Model Comparison to compare intelligence, coding signal, speed, and API value; use the AI subscription price comparison to compare regional storefront prices; and use the AI Plan Finder when you want a shortlist based on budget and work style. These tools answer different questions, so they link to one another without merging price, capability, access, and troubleshooting into one opaque score.
Claude: choose the work surface before the plan
A chat window, an Office add-in and an API client can use Claude without being the same purchase or workflow. First identify where the result needs to land; then check the account, rollout and billing conditions. This is a route guide, not a quality ranking.
| What you need | Start here | Boundary to check |
|---|---|---|
| Delegate a report, research task or finished-file workflow. | Claude conversation / unified Cowork experience. | Gradual Pro/Max rollout. Cloud work can continue after leaving; local files or apps still need Desktop open. |
| Edit an open workbook, deck, document or email. | The appropriate Claude for Microsoft 365 add-in. | Paid Claude plan, supported Office host and deployment settings. Cross-app mode requires direct Claude sign-in, not a gateway route. |
| Work interactively on a code repository. | Claude Code with the intended sign-in. | Check subscription versus Console authentication; the included Max API grant does not cover interactive Code. |
| Run your own application or agent. | Claude API / Console organization. | Confirm the key’s organization, eligible credit coverage, spend cap and expiry. App usage and Console balance remain different ledgers. |
- Still seeing Chat and Cowork choices? That can mean the unified rollout has not reached your account, not that another upgrade will fix it. Follow the official rollout check.
- Office is not the chat model picker. Check the supported Excel builds and curated model list. Outlook is beta, needs Exchange Online and Microsoft Graph admin consent; do not bypass your organization’s controls.
- Before enabling cross-app work: read the per-add-in switches and organization rules. Use a trusted file copy, review edits and confirm what is open before authorizing consequential actions.
For a purchase decision, compare Pro/Max capacity and usable API-credit value. For a billing mismatch, use the funded-but-blocked checklist or Code sign-in checks. The API-credit program does not turn every Claude surface into one prepaid wallet.
Claude work-surface conditions checked October 19, 2026 against the linked official guides. Other model cards keep their own review scopes. No private account, add-in installation, model output or payment was tested.
Access signals are not unlimited usage
“Included” or “listed” means the provider currently documents or exposes a model or feature at that plan level. Rollouts, country, account type, workspace policy and platform can change what appears in the live model picker.
A paid plan can still have message, context, feature or capacity limits, and consumer access is separate from API billing. Treat each card as a source-linked snapshot and confirm the provider’s current plan page before purchasing.
Need software around the model, not only plan access? Compare deployment choices in the Self-Hosted AI Tools directory, then confirm the provider and model terms separately.
New access checks, October 1: compare Sonnet 5.5 and Opus 5.5, or read the DevDay guide to GPT-6.1 Sol, Dots and Astra Ultrafast eligibility. These are different products and access routes, not interchangeable subscription benefits.