Tool detail

ChatGPT

ChatGPT is the lowest-risk default when one subscription needs to span research, writing, meetings, ChatGPT Work deliverables, and Codex-backed workflow work instead of only the IDE. The operational distinction matters: Chat is conversational, Work is outcome-oriented, and Codex is for software development.

Check current pricing

ChatGPT: See pricing, plans, and team cost Compare top alternative: Claude

Updated because: Do not replace a production GPT-5.6 budget with Astra assumptions yet. Treat Astra as a separately governed high-cost pilot until access, safety controls, and workload gains are validated for the buyer's organization.

Best for: Individual · Research • 10/10
Avoid if: Coding is better than general assistants used to be, but still not as IDE-native as Cursor.
Starting price: $20 /mo

Last verified: Sep 4, 2026

Some links on AgentHub may be affiliate or partner links. We may earn a commission at no extra cost to you. Learn more

Choosing a tool

Strengths and limits

Check the reasons to choose it and the limitations.

Choosing a tool

GPT-6 Astra materially raises the top-end capability and price ceiling, but it is not yet a generally available default. Buyers should separate a Trusted Access pilot from current GPT-5.6 production budgeting.
ChatGPT now has a clearer self-serve ladder from Free to Go to Plus, which lowers the jump from casual use to a paid general-purpose AI seat.
ChatGPT Pro now has $100 and $200 monthly tiers, so heavy individual users can buy 5x or 20x more Plus-level usage without moving into a team workspace.
Bundling connectors and Codex into the same workspace lowers the need to buy a separate chat seat plus coding seat for every mixed-role team member.
Role-specific Codex plugins, annotations, and Sites shift Codex from a coding-only helper toward a workflow surface beyond coding for analysts, marketers, operators, and product teams.
At $20 per user per month on annual billing, ChatGPT Business is now materially cheaper than its $30 monthly path. The deciding variable for coding-heavy teams is no longer a fixed Codex-only seat: it is included allowance plus token-based credit consumption.
ChatGPT Work creates a third operating surface between Chat and Codex. It expands the case for a mixed-role workspace, but its Codex-aligned consumption, app permissions, and desktop local-access boundary mean it should be budgeted and governed as an agent, not marketed as unlimited chat.

Trade-offs

Coding is better than general assistants used to be, but still not as IDE-native as Cursor.
Enterprise pricing remains sales-led, so large-scale cost planning is less transparent.
The product surface is broad enough that teams need governance to avoid tool sprawl inside one workspace.
Work availability and access differ by surface, plan, workspace, and enabled apps; do not promise local-file or write access before validating desktop permissions and admin action controls.

Fit by user

Who is it for?

Review the trade-offs for individuals, teams, and enterprises.

Individual

10/10

Best use case: Research

For solo operators, ChatGPT is the widest one-seat answer when coding is only one part of the day and research or writing quality matters just as much.

Coding 9/10 · Research 10/10 · Meetings 7/10 · Automation 8/10 · Writing 9/10 · Customer service 6/10

Team

9/10

Best use case: Research

For teams, Business closes much of the gap between personal chat subscriptions and a governed workspace by combining admin controls, shared context, and connectors.

Coding 8/10 · Research 9/10 · Meetings 8/10 · Automation 8/10 · Writing 8/10 · Customer service 7/10

Enterprise

8/10

Best use case: Research

For enterprises, ChatGPT is strongest when leaders want one cross-functional assistant, but final enterprise buying still needs a sales conversation for final pricing and controls.

Coding 7/10 · Research 8/10 · Meetings 8/10 · Automation 8/10 · Writing 8/10 · Customer service 7/10

Individual plans

For individual subscribers

Check the features and costs for your own work.

For solo operators, ChatGPT is the widest one-seat answer when coding is only one part of the day and research or writing quality matters just as much.

Pricing

Plans and pricing

Compare plan prices, features, and usage limits.

Free plan available · Paid from $20/mo

Free

$0 / month

$0 per seat / month on annual billing

Included models: GPT-5.5 Instant, GPT-5.6 Terra

Limited GPT-5.5 Instant access, with limited GPT-5.6 Terra access in Codex and ChatGPT Work
Limited messages, uploads, and image generation
Limited deep research and memory

Verification status: Review required · 31 days since verification
Last checked: Sep 4, 2026
Scope: Global English availability · Plan: Individual
Official evidence: ChatGPT Free pricing

Official pricing

Go

Custom quote

No annual price published

Included models: GPT-5.5 Instant, GPT-5.6 Terra

More GPT-5.5 Instant access plus limited GPT-5.6 Terra access in Codex and ChatGPT Work
More messages, uploads, and image creation
Longer memory and past-chat recall
Data analysis, search, canvas, and projects

Verification status: Review required · 31 days since verification
Last checked: Sep 4, 2026
Scope: Global English availability · Plan: Individual
Official evidence: ChatGPT Go pricing

Official pricing

Plus

Popular

$20 / month

No annual price published

Included models: GPT-5.5 Instant, GPT-5.6 Sol, GPT-5.6 Terra, GPT-5.6 Luna

GPT-5.6 Sol, Terra, and Luna with configurable reasoning effort
Expanded messages and uploads
Expanded and faster image creation
Expanded deep research and agent mode

Verification status: Review required · 31 days since verification
Last checked: Sep 4, 2026
Scope: Global English availability · Plan: Individual
Official evidence: ChatGPT Plus pricing

Official pricing

Pro 5x

$100 / month

No annual price published

5x more usage than Plus
GPT-5.6 Sol Pro and advanced reasoning access
Expanded file uploads, deep research, agent mode, memory, and context
ChatGPT Work for longer multi-step deliverables

Verification status: Review required · 31 days since verification
Last checked: Sep 4, 2026
Scope: Global English availability · Plan: Individual
Official evidence: ChatGPT Pro 5x pricing

Official pricing

Pro 20x

$200 / month

No annual price published

20x more usage than Plus
Highest ChatGPT Pro usage tier
Maximum file uploads, deep research, agent mode, memory, and context
ChatGPT Work for longer multi-step deliverables

Verification status: Review required · 31 days since verification
Last checked: Sep 4, 2026
Scope: Global English availability · Plan: Individual
Official evidence: ChatGPT Pro 20x pricing

Official pricing

Business

$30 / month

$20 per seat / month on annual billing

Included models: GPT-5.5 Instant, GPT-5.6 Sol, GPT-5.6 Terra, GPT-5.6 Luna

Standard Business seats include a Codex usage allowance
Most Codex usage and purchased credits now follow a token-based rate card
ChatGPT Work for multi-step documents, analyses, presentations, and Sites across enabled apps and files
New Business workspaces cannot add Codex-only pay-as-you-go seats; existing Codex-only seats are unaffected

Verification status: Review required · 31 days since verification
Last checked: Sep 4, 2026
Scope: Global English availability · Plan: Team
Official evidence: ChatGPT Business pricing · ChatGPT pricing

Official pricing

Enterprise

Custom quote

No annual price published

Included models: GPT-5.5 Instant, GPT-5.6 Sol, GPT-5.6 Terra, GPT-5.6 Luna

Everything in Business with expanded context windows and larger files
ChatGPT Work access is governed by workspace, role, and Enterprise Key Management eligibility
SCIM, EKM, analytics, domain verification, and role-based access controls
Custom data retention policies and data residency in ten regions

Verification status: Review required · 31 days since verification
Last checked: Sep 4, 2026
Scope: Global English availability · Plan: Enterprise
Official evidence: ChatGPT Enterprise pricing

Official pricing

FAQ

Common questions

Answers about pricing and usage terms.

Does ChatGPT Business include coding workflows?

Yes. Standard Business seats include a Codex allowance. For most migrated workspaces, credits now map to input, cached-input, and output-token use; new Business workspaces cannot add Codex-only pay-as-you-go seats, while existing ones are unaffected. Codex is still a general-workspace coding surface rather than an IDE-native cockpit.

How should a team budget Codex inside ChatGPT Business?

Treat the annual Business seat and heavy agent consumption as separate questions. Start with the included allowance, then use the token-based rate card to model representative input, cached-input, and output mix. Do not assume a new workspace can solve overage with a Codex-only pay-as-you-go seat.

When is ChatGPT Go worth considering?

Go matters when Free is too constrained but Plus is still more capability than the buyer needs. It mainly buys more GPT-5.5 Instant access, more uploads and images, longer memory, and limited GPT-5.6 Terra access in Codex and ChatGPT Work.

How should a team choose between ChatGPT Chat, Work, and Codex?

Use Chat for quick questions and iteration, Work when the requested outcome is a multi-step document, analysis, spreadsheet, presentation, report, or Site, and Codex for code, repositories, tests, and developer tools. Work follows the Codex usage structure, so a long deliverable is not automatically covered by the visible seat price.

What must an admin validate before enabling ChatGPT Work?

Validate the plan and workspace eligibility, role access, enabled apps and their allowed actions, plus whether the pilot needs cloud Work or desktop access to local files and apps. For file creation, confirm the target surface: Google Docs, Sheets, and Slides need the corresponding enabled Google app; direct Excel control is a Codex desktop add-in flow, not a Work guarantee.

When should a team buy Business instead of handing out Plus seats?

Buy Business when shared governance, SAML SSO, connectors, shared projects, and business-data controls matter. Plus remains fine for loose individual usage, but it is not a governed team rollout surface.

Does OpenAI train on ChatGPT Business or Enterprise data?

Business and Enterprise are positioned with no training on business data by default. That is one of the main reasons to move off personal plans when rollout becomes company-wide.

Is ChatGPT still the best buy if coding is the main job?

Usually only if the same seat also needs research, writing, and cross-tool work. For a pure IDE-native coding cockpit, Cursor or GitHub Copilot can still be better buys.

Alternatives

When to consider an alternative to ChatGPT

Keep ChatGPT when one seat still has to cover research, writing, meetings, and coding-adjacent work together. Switch only when the seat exists for one dominant workflow: Claude for reasoning-heavy writing, Perplexity for citation-backed research, Gemini for Google Workspace rollout.

Claude

Claude is the cleanest alternative when the team keeps reaching for ChatGPT mainly for difficult reasoning or editorial work rather than breadth.

Perplexity

Perplexity is the better alternative when the daily job is finding cited answers quickly, not maintaining a broad assistant workspace.

Gemini

Gemini becomes the practical alternative when Google Workspace rollout and admin familiarity outweigh the value of ChatGPT's wider product surface.

Watchlist

Track changes for this tool

Save without an account, review price and limit changes, and choose email alerts from your watchlist.

Save this shortlist

Model details

Models, benchmarks, and API pricing

Review supported models and published measurements.

Models

GPT-6 Astra

Context window: 1,050,000 tokens
Max output: 128,000 tokens
Training cutoff: Apr 30, 2026

GPT-5.5 Instant

Context window: 128,000 tokens

Available in: Free, Go, Plus, Pro, Business, Enterprise

GPT-5.6 Sol

Context window: 1,050,000 tokens
Max output: 128,000 tokens
Training cutoff: Feb 16, 2026

Available in: Plus, Pro, Business, Enterprise

GPT-5.6 Terra

Context window: 1,050,000 tokens
Max output: 128,000 tokens
Training cutoff: Feb 16, 2026

Available in: Free, Go, Plus, Pro, Business, Enterprise

GPT-5.6 Luna

Context window: 1,050,000 tokens
Max output: 128,000 tokens
Training cutoff: Feb 16, 2026

Available in: Plus, Pro, Business, Enterprise

Benchmarks

Benchmarks
Dimension	Score	Model	Measured	Source
Terminal-Bench 2.1	88.8%	GPT-5.6 Sol	Jul 9, 2026	Source
SWE-Bench Pro (Public)	64.6%	GPT-5.6 Sol	Jul 9, 2026	Source
OSWorld 2.0	62.6%	GPT-5.6 Sol	Jul 9, 2026	Source
Toolathlon	58%	GPT-5.6 Sol	Jul 9, 2026	Source
Agents' Last Exam	52.7%	GPT-5.6 Sol	Jul 9, 2026	Source

API pricing

API pricing
Model	Input (per 1M tokens)	Output (per 1M tokens)	Documentation
GPT-6 Astra	$10	$50	Frontier model for the hardest end-to-end work. Access is initially limited to Trusted Access enterprises, with broader API and plan rollout announced for the coming days.
GPT-5.6 Sol	$4	$20	Flagship GPT-5.6 tier for coding, professional work, and long-running agents.
GPT-5.6 Terra	$2	$12	Balanced GPT-5.6 tier for cost-sensitive professional and agentic workloads.
GPT-5.6 Luna	$0.20	$1.20	Fastest, lowest-cost GPT-5.6 tier for scaled routines and subagents.
o4-mini	$1.10	$4.40	Legacy small reasoning model, now succeeded by GPT-5 mini.

Changes

Recent changes

Do not replace a production GPT-5.6 budget with Astra assumptions yet. Treat Astra as a separately governed high-cost pilot until access, safety controls, and workload gains are validated for the buyer's organization.

Sep 4, 2026high

GPT-6 Astra enters limited rollout above the GPT-5.6 family.

OpenAI introduced GPT-6 Astra with a 1.05M context window, 128K maximum output, and $10/$50 per million input/output token pricing. Initial access is limited to Trusted Access enterprises; broader API and Plus, Pro, Business, and Enterprise access was announced for the coming days.

Open tool change history

chatgpt-vs-claude chatgpt-vs-gemini

Sep 4, 2026medium

GPT-5.6 API prices fall across Sol, Terra, and Luna.

Current OpenAI model documentation lists Sol at $4/$20, Terra at $2/$12, and Luna at $0.20/$1.20 per million input/output tokens.

Re-run routing and unit-cost models: Luna is now one-fifth of its launch input and output price, while Sol and Terra also became cheaper for production agent workloads.

Open tool change history

chatgpt-vs-claude chatgpt-vs-gemini

Aug 14, 2026high

ChatGPT Work creates a distinct outcome-agent surface between Chat and Codex.

ChatGPT Work handles longer multi-step deliverables across enabled apps and files, including documents, spreadsheets, presentations, reports, and Sites. Cloud Work is available through web and mobile, while desktop access to local files and apps needs permission. Its usage follows the same structure as Codex rather than an unlimited-chat model.

Budget Work as an agentic consumption surface and validate workspace eligibility, roles, enabled apps, allowed actions, and cloud-versus-local access before rollout. Do not assume an included chat seat grants every file-creation or desktop-control workflow.

Open tool change history

chatgpt-vs-claude

Aug 13, 2026high

ChatGPT Business annual pricing is now $20 per user per month, while Codex usage moves to a token ratecard.

OpenAI reduced annual Business pricing from $25 to $20 per user per month while monthly pricing remains $30. Codex usage and purchased credits now follow the token ratecard, and new Codex-only pay-as-you-go seats are no longer available for Business workspaces.

Business becomes a lower fixed-seat baseline, but coding-heavy teams must separately forecast token-credit consumption. Do not use the Business seat price as the complete Codex budget.

Open tool change history

chatgpt-vs-claude chatgpt-vs-grok chatgpt-vs-perplexity

Jul 11, 2026high

GPT-5.6 replaces GPT-5.5 across ChatGPT, Codex, and the OpenAI API.

OpenAI launched GPT-5.6 Sol, Terra, and Luna on July 9 with configurable reasoning, multi-agent and programmatic tool-calling paths, and API prices of $5/$30, $2.50/$15, and $1/$6 per million input/output tokens.

Teams should now model OpenAI as a three-tier GPT-5.6 family instead of a single premium flagship. Sol keeps GPT-5.5's flagship token price while improving agentic and computer-use performance; Terra and Luna create clearer cost-down paths for scaled assistants and subagents.

Open tool change history

chatgpt-vs-claude chatgpt-vs-gemini chatgpt-vs-grok

Jul 3, 2026medium

ChatGPT Pro is now split into $100 and $200 monthly tiers.

Heavy individual buyers can now choose a lower Pro entry point before jumping to the highest usage tier.

Open tool change history

chatgpt-vs-perplexity customgpt-vs-chatgpt glean-vs-chatgpt

Next reads

Compare with other tools

Compare prices and features with similar tools.

ChatGPT vs Claude

Claude

ChatGPT is the lower-friction mixed-workload default, especially for cloud and app-connected Work deliverables. Claude is the better pick when reasoning quality, long-form output, or controlled work in a selected local workspace outweigh ecosystem breadth.

ChatGPT vs Gemini

Gemini

ChatGPT is the better broad default when one AI seat has to cover many kinds of work. Gemini is the better buy when the team already runs on Google Workspace and wants AI bundled into docs, meetings, search, and NotebookLM.

ChatGPT vs Grok

Grok

ChatGPT remains the lower-risk company-wide default, while Grok is now a credible specialist challenger for a bounded automation, voice-bot, Build Mode, or coding-agent pilot—not a single interchangeable chat subscription.

Next reads

Compare with other tools

Compare prices and features with similar tools.

Pricing guideChatGPT pricing guideChatGPT still has Free, Go, Plus at $20/month, and Business at $20/user/month annually or $30 monthly. Work follows the same usage structure as Codex, so model longer Work deliverables and Codex separately from the Business seat: allowance and additional credits vary by plan and task. Pro remains $100/month for 5x Plus usage or $200/month for 20x, with no annual billing for Go, Plus, or Pro.Alternatives guideRead alternatives guideKeep ChatGPT when one seat still has to cover research, writing, meetings, and coding-adjacent work together. Switch only when the seat exists for one dominant workflow: Claude for reasoning-heavy writing, Perplexity for citation-backed research, Gemini for Google Workspace rollout.Use casesAI coding tools for solo developers: shortlist and fit guideFor solo developers, indie hackers, and technical operators choosing one paid AI seat they will actually open every day, with clear limits on agent access, review, rollback, and spend.ChangesSee ChatGPT change historyDo not replace a production GPT-5.6 budget with Astra assumptions yet. Treat Astra as a separately governed high-cost pilot until access, safety controls, and workload gains are validated for the buyer's organization.

Features

Features by plan

Check which features require a paid plan.

AI · 3

Deep research

Turns open-ended questions into multi-step research with citations and synthesis.

Available in: Free, Plus, Pro, Business, Enterprise

Verification status: Review required · 31 days since verification
Last checked: Sep 4, 2026
Scope: Global English availability · Plan: Free, Plus, Pro, Business, Enterprise
Official evidence: ChatGPT pricing

ChatGPT Work

An outcome-oriented agent for longer, multi-step work: it can research, analyze sources, and create documents, spreadsheets, presentations, reports, or Sites across enabled apps and files. Use Chat for fast conversation, Work for a deliverable, and Codex for repository and software tasks.

Available in: Free, Go, Plus, Pro, Business, Enterprise

Verification status: Review required · 31 days since verification
Last checked: Sep 4, 2026
Scope: Global English availability · Plan: Free, Go, Plus, Pro, Business, Enterprise
Official evidence: ChatGPT Work and Codex · Creating deliverables with ChatGPT Work

Codex agent

Lets ChatGPT reason across codebases, related documentation, and role-specific Codex plugins from the same workspace; Business buyers should model the included allowance and any token-based credits separately.

Available in: Plus, Pro, Business, Enterprise

Verification status: Review required · 31 days since verification
Last checked: Sep 4, 2026
Scope: Global English availability · Plan: Plus, Pro, Business, Enterprise
Official evidence: ChatGPT pricing · Codex rate card

Integration · 2

Apps and connectors

Connects internal tools and cloud storage for grounded answers inside the workspace.

Available in: Business, Enterprise

Verification status: Review required · 31 days since verification
Last checked: Sep 4, 2026
Scope: Global English availability · Plan: Business, Enterprise
Official evidence: Apps in ChatGPT

Codex plugins and Sites

Packages apps, skills, instructions, workflows, and shareable Sites so teams can standardize Codex beyond software-only tasks.

Available in: Business, Enterprise

Verification status: Review required · 31 days since verification
Last checked: Sep 4, 2026
Scope: Global English availability · Plan: Business, Enterprise
Official evidence: Codex plugins documentation · ChatGPT pricing

Security · 3

Work surfaces and permission boundaries

Work on web and mobile runs in the cloud; desktop Work can use local files and desktop apps only with user permission. Access, enabled apps, and actions remain constrained by the plan, workspace, admin policy, and role, so a deliverable agent should be piloted with explicit read/write boundaries.

Available in: Free, Go, Plus, Pro, Business, Enterprise

Verification status: Review required · 31 days since verification
Last checked: Sep 4, 2026
Scope: Global English availability · Plan: Free, Go, Plus, Pro, Business, Enterprise
Official evidence: ChatGPT Work and Codex · Apps in ChatGPT

Codex usage and credit controls

Makes Codex consumption visible as token-based credits for most migrated plans, so heavy agent work can be budgeted by input, cached-input, and output-token mix instead of a simple request count.

Available in: Plus, Pro, Business, Enterprise

Verification status: Review required · 31 days since verification
Last checked: Sep 4, 2026
Scope: Global English availability · Plan: Plus, Pro, Business, Enterprise
Official evidence: Codex rate card

Business admin controls

Adds workspace controls, SAML SSO, and no-training defaults for business data.

Available in: Business, Enterprise

Verification status: Review required · 31 days since verification
Last checked: Sep 4, 2026
Scope: Global English availability · Plan: Business, Enterprise
Official evidence: ChatGPT Business · Apps in ChatGPT

Collaboration · 2

Projects and shared projects

Keeps recurring work organized in project spaces before a team needs a fully governed business workspace.

Available in: Go, Plus, Pro, Business, Enterprise

Verification status: Review required · 31 days since verification
Last checked: Sep 4, 2026
Scope: Global English availability · Plan: Go, Plus, Pro, Business, Enterprise
Official evidence: ChatGPT pricing

Custom GPTs

Packages prompts, tools, files, and tasks into reusable workflows once teams move above the Go tier.

Available in: Plus, Pro, Business, Enterprise

Verification status: Review required · 31 days since verification
Last checked: Sep 4, 2026
Scope: Global English availability · Plan: Plus, Pro, Business, Enterprise
Official evidence: ChatGPT pricing

Rollout policy evidence

Check policy and admin conditions against official evidence

These rollout signals come from security, help, and product documentation. Keep legal and security approval in your organization’s formal review process.

Verification status: Review required · 31 days since verification · Plan-specific conditions need confirmation
Last checked: Sep 4, 2026
Scope: Global English availability · Plan: Confirm official terms per plan
Official evidence: GPT-6 Astra model documentation · ChatGPT release notes · GPT-6 Astra safety overview · GPT-5.6 Sol model documentation · GPT-5.6 Luna model documentation · OpenAI model catalog · OpenAI API pricing · Codex plugins documentation · ChatGPT Pro tiers · Codex rate card · ChatGPT Work and Codex · Creating deliverables with ChatGPT Work · Apps in ChatGPT

Best lists

Shortlists by use case

Explore other tools for the same tasks.

Best AI meeting assistants by suite and follow-through

Use this list to choose an AI meeting assistant around the system your team already uses. It weighs suite alignment, capture quality, and whether action items stay in the same workflow after the call.

Best AI research assistants for sourced decision-making

Choose an AI research assistant by its evidence boundary: cited live-web discovery, synthesis from a known source pack, search across connected company apps, or a general research seat that continues into planning and writing. The ranking rewards answers a buyer can inspect and operationalize.

Best AI writing tools for real team workflows

Use this shortlist when the writing seat could mean careful drafting, mixed-workload support beyond drafting, or workspace-native publishing. The ranking keeps review loops, research spillover, and rollout overhead in the same buying conversation.

Best AI coding assistants by workflow

Choose an AI coding assistant by the work it must own: parallel agent execution, governed review, local-to-cloud handoff, or a mixed seat that also handles research and writing. This shortlist compares those operating models before headline model claims or seat price.

Best enterprise AI tools when governance changes the decision

Use this list when the purchase is an enterprise AI rollout, not a single-user assistant choice. It weighs permissions-aware retrieval, admin control, and whether the rollout needs to stay inside an existing enterprise suite standard.

Best AI assistants for small product teams

Use this list when a small product team needs one AI seat to cover planning, docs, meetings, light technical work, and mixed-role collaboration. The ranking favors tools that can survive a real budget conversation without forcing several narrow subscriptions.