Multimodal Foundation
GPT-4o
Text and image understanding.
One Model Family · One Curated Home
ChatGPT on GlobalGPT
Choose GPT-6 Astra for demanding work, GPT-6.1 Sol for coding, GPT-6.1 Terra for visual reasoning and writing, or GPT-6 Luna for focused tasks. Create and edit images with GPT Image 2.5.
Curated by GlobalGPT Review Team
A focused home for the ChatGPT series inside GlobalGPT—not another all-model marketplace.
5Current Picks
10Archived Model Entries
7Guide Topics
A Selected Path Through the ChatGPT Series.
Selected text and image milestones, with Sol and Terra as the current GPT-6.1 additions.
Earlier generations
GPT-5.5 & 5.6
Previous GPT models for professional work.
September 2026
GPT-6 Astra
Demanding reasoning and coding.
September 2026
GPT Image 2.5
Image generation and editing.
Five Current Choices for Different Tasks.
Astra for demanding work, Sol for coding, Terra for visual reasoning and writing, Luna for focused tasks, and GPT Image 2.5 for image generation and editing.
01 / 05
Text + Image Input
GPT-6 Astra
A starting point for complex reasoning, demanding coding, research and document creation. Use the model in GlobalGPT’s chat workspace.
Complex ReasoningCodingProfessional Work
Image Generation
GPT Image 2.5
Generate and edit images from text and image inputs. GlobalGPT provides one GPT Image 2.5 entry: describe your creative brief or precise change in the prompt.
GenerateEditVisual Layout
Balanced
GPT-6.1 Sol
A balanced choice for complex coding and professional work. OpenAI positions it near Astra in performance at a lower API cost; choose the route that fits your workflow.
CodingProfessional WorkBalanced
Visual reasoning & writing
GPT-6.1 Terra
A balanced option for visual reasoning, natural writing and computer-use workflows. OpenAI documents text and image inputs with text output; available tools depend on the GlobalGPT workspace.
Visual ReasoningWritingBalanced
High Volume
GPT-6 Luna
An efficient choice for focused, high-volume tasks such as extraction, classification, concise summaries and everyday questions.
Focused TasksHigh Volume
Previous Flagships & Earlier ModelsGPT-6 Sol, the GPT-5.6 family, GPT Image 2 and earlier
GPT-6 SolThe earlier Sol model for coding and agentic workflows. GPT-6.1 Sol is the newer balanced choice./home/gpt-6-sol?inviter=hub_research_chatgpt&login=1
GPT-5.6 TerraFaster & More Efficient · Balanced · Released Jul 2026An earlier balanced GPT-5.6 model, retained for existing workflows.Daily Work · Writing · Balanced/home/gpt-5-6-terra?inviter=hub_research_chatgpt&login=1Official source
GPT-5.6 LunaFaster & More Efficient · High Volume · Released Jul 2026The earlier GPT-5.6 efficient model, retained for existing workflows.Fast Tasks · High Volume/home/gpt-5-6-luna?inviter=hub_research_chatgpt&login=1Official source
GPT-5.6 SolAn earlier flagship that remains available for focused drafting, research and coding tasks./home/gpt-5-6-sol?inviter=hub_research_chatgpt&login=1
GPT Image 2The earlier image-generation and editing entry, retained for existing workflows./image-generator/gpt-image-2?inviter=hub_research_chatgpt&login=1
GPT-5.5Previous flagship for difficult coding and professional workloads · Released Apr 2026./home/gpt-5-5?inviter=hub_research_chatgpt&login=1
GPT-5.4Previous flagship for professional and coding work./home/gpt-5-4?inviter=hub_research_chatgpt&login=1
GPT-5.3Earlier ChatGPT generation retained for reference./home/gpt-5-3?inviter=hub_research_chatgpt&login=1
GPT-5.2Previous flagship for professional work./home/gpt-5-2?inviter=hub_research_chatgpt&login=1
GPT-4oEarlier fast, flexible multimodal GPT model./home/gpt-4o?inviter=hub_research_chatgpt&login=1
Compare Models, Access and Costs
Choose by workload, then review official subscriptions, developer API prices and published benchmark conditions.
OpenAI API Token Prices
| Model / API | Input / 1M tokens | Cached input / 1M | Output / 1M | Billing conditions |
|---|---|---|---|---|
| GPT-6 AstraOpenAI OpenAI official documentation ↗Last verified: | $10.00 | $1.00 | $50.00 | Standard speed; prompt ≤272K tokens |
| GPT-6.1 SolOpenAI OpenAI official documentation ↗Last verified: | $2.00 | $0.10 | $10.00 | Standard speed; prompt ≤272K tokens |
| GPT-6.1 TerraOpenAI OpenAI official documentation ↗Last verified: | $1.00 | $0.05 | $5.00 | Standard speed; prompt ≤272K tokens |
| GPT-6 LunaOpenAI OpenAI official documentation ↗Last verified: | $0.10 | $0.01 | $0.50 | Standard speed; prompt ≤272K tokens |
USD. Standard text-token rates for prompts up to 272K. Above 272K, the full request has 2× input/cached-input and 1.5× output rates. Cache writes, tools, other speed tiers and regional processing can add charges. These are direct OpenAI API rates, not GlobalGPT subscription prices. Image tokens use separate rates.
Official provider documentation ↗Last verified:
Official ChatGPT Plan Prices
| Plan | USD monthly | Billing | Work / Codex scope in the cited plan guide |
|---|---|---|---|
| FreeOpenAI OpenAI official documentation ↗Last verified: | $0 | Free plan | GPT-6 Luna desktop rollout; usage limits |
| GoOpenAI OpenAI official documentation ↗Last verified: | $8 | Monthly | Lightweight Codex tasks; GPT-6 Luna rollout |
| PlusOpenAI OpenAI official documentation ↗Last verified: | $20 | Monthly | Sol and Luna; Codex tools and usage allowances |
| ProOpenAI OpenAI official documentation ↗Last verified: | $100 / $200 / $500 | Monthly plan options | Higher usage options; Ultrafast on the $500 tier |
| BusinessOpenAI OpenAI official documentation ↗Last verified: | $20 annual billing / $25 monthly billing per user | Minimum 2 users | Managed team workspace and usage controls |
Official ChatGPT subscription prices in OpenAI’s Work/Codex guide. ChatGPT Work and Codex share usage; listed model scope refers to that guide, not every ChatGPT surface. API token billing and GlobalGPT plans are separate. Local taxes and availability may differ.
Official provider documentation ↗Last verified:
GPT Model Benchmarks Compared at Max Reasoning Effort
| GPT model | Professional tasks: Agents’ Last Exam v1 | Visual reasoning: MMMU-Pro | Computer use: OSWorld 2.1 offline | Factual error rate ↓ |
|---|---|---|---|---|
| GPT-6 AstraOpenAIAll four results use Max reasoning effort in this source. OSWorld 2.1 uses the offline set. The factuality test uses deliberately error-inducing prompts and does not measure typical-use error rates. OpenAI GPT-6.1 Terra launch evaluation charts ↗Last verified: | 59.3% | 86.7% | 80.9% | 4.2% |
| GPT-6.1 SolOpenAIAll four results use Max reasoning effort in this source. OSWorld 2.1 uses the offline set. The factuality test uses deliberately error-inducing prompts and does not measure typical-use error rates. OpenAI GPT-6.1 Terra launch evaluation charts ↗Last verified: | 58.6% | 86.3% | 78.0% | 4.4% |
| GPT-6.1 TerraOpenAIAll four results use Max reasoning effort in this source. OSWorld 2.1 uses the offline set. The factuality test uses deliberately error-inducing prompts and does not measure typical-use error rates. OpenAI GPT-6.1 Terra launch evaluation charts ↗Last verified: | 60.0% | 85.5% | 74.2% | 4.3% |
| GPT-6 LunaOpenAIAll four results use Max reasoning effort in this source. OSWorld 2.1 uses the offline set. The factuality test uses deliberately error-inducing prompts and does not measure typical-use error rates. OpenAI GPT-6.1 Terra launch evaluation charts ↗Last verified: | 50.9% | 79.9% | 55.5% | 18.4% |
All four GPT models are compared at Max reasoning effort using the same set of OpenAI GPT-6.1 Terra launch charts. Higher is better for the first three tests; lower is better for factual error rate. Factuality uses deliberately difficult prompts selected after users flagged earlier errors, so it is not a typical-use error rate. OpenAI research/API results depend on the evaluation harness and are not GlobalGPT measurements. These task-specific results do not establish an overall model ranking.
OpenAI GPT-6.1 Terra launch evaluation charts ↗Last verified:
| Model / product | Workload | Inputs and output | Access in GlobalGPT |
|---|---|---|---|
| GPT-6 AstraOpenAI OpenAI official documentation ↗Last verified: | Demanding reasoning and professional work | Text / image input; text output | Chat model entry |
| GPT-6.1 SolOpenAI OpenAI official documentation ↗Last verified: | Coding and professional work | Text / image input; text output | Chat model entry |
| GPT-6.1 TerraOpenAI OpenAI official documentation ↗Last verified: | Visual reasoning and natural writing | Text / image input; text output | Chat model entry |
| GPT-6 LunaOpenAI OpenAI official documentation ↗Last verified: | Focused, high-volume tasks | Text / image input; text output | Chat model entry |
| GPT Image 2.5OpenAIGlobalGPT has one 2.5 entry, not separate Sunburst and Flare selectors. OpenAI official documentation ↗Last verified: | Image generation and editing | Text / image input; image output | Dedicated image workspace |
Task fit follows the provider’s positioning and the visible GlobalGPT route. This is not a measured speed or quality ranking.
Browse by the Question You Have.
Each guide has one primary home. Model tags connect related reading without duplicating the same article across the page.
Access & SetupGlobalGPT vs ChatGPT: What’s Different?
Separate model access from app features, and compare the two platforms’ workflows and subscriptions.
Platforms & AccessGlobalGPTChatGPT vs Playground vs GPT Store
Platforms & Access01GlobalGPTChatGPT Subscription Plans in 2026
Plans & Access02GlobalGPTHow to Use ChatGPT
12
Choose an app, API testing workspace or custom assistant without confusing subscriptions with API credits.
Platforms & Access01GlobalGPTChatGPT Subscription Plans in 2026Compare model access, plan limits and the differences between individual and team subscriptions.
Plans & Access02GlobalGPTHow to Use ChatGPTChoose a workspace and model, write a clear first prompt, and use files, images or coding tools for the task.
Getting StartedGuide03GlobalGPTHow to Use ChatGPT EffectivelyStructure instructions, provide context and improve a draft through focused follow-up prompts.
Getting StartedWorkflow04GlobalGPTWhy Is ChatGPT Not Working?Diagnose service incidents, account issues, browser problems and tool failures before retrying.
Limits & TroubleshootingTroubleshooting05GlobalGPTWhy Is ChatGPT So Slow?Distinguish service delays from reasoning, long conversations, tool work and local network problems.
Limits & TroubleshootingTroubleshooting06GlobalGPTHow to Cancel ChatGPT PlusCancel through the web, Apple or Google billing channel that originally handled your subscription.
Limits & TroubleshootingSetup07GlobalGPTHow to Use ChatGPT on iPhoneExplore the official iPhone app and Apple integration, with account and device requirements kept separate.
Devices & AssistantsMobile08GlobalGPTHow to Use ChatGPT With SiriConnect Siri through supported Apple integration or Shortcuts and troubleshoot the access route.
Devices & AssistantsIntegration09GlobalGPTChatGPT Plus Free Trial: Eligibility & OffersUnderstand offer eligibility and the difference between free access, trials and paid subscriptions.
Plans & OffersAccess10GlobalGPTChatGPT Plus vs BusinessCompare an individual subscription with company administration, shared workspaces and business controls.
Plans & OffersPlans11GlobalGPTChatGPT Plus Limits: Caps, Reset Times and What to Do NextUnderstand Plus usage limits, rolling windows and what happens when an allowance is reached.
12Start Closer to a Useful Answer.
Prompt collections and practical workflows from the GlobalGPT Hub, organized around what you want ChatGPT to help you make.
Writing
Turn rough intent into clearer professional communication.
Write Better EmailsBuild a Stronger ResumeImage
Use repeatable structures for generation, editing and visual direction.
Best GPT Image 2 PromptsChatGPT Prompts for ImagesGPT Image 2.5: Workflow & Editing ReviewResearch
Frame questions, gather evidence and organize a deeper investigation.
Use ChatGPT Deep ResearchPrompts for Better AccuracyCoding
Structure coding requests and continue the work in a focused canvas.
Use ChatGPT CanvasChoose a Coding ModelWhat Makes a GPT Output Ready to Use?
Original research on how people describe the work they want GPT to do—and the control they want to retain.
What Makes a GPT Output Usable? Task Delegation and User Control in GlobalGPT Requests
Published:
Beyond a relevant answer, what makes an AI-generated result ready to use?
A 28-day GlobalGPT study combines platform activity statistics with qualitative analysis of selected requests. It examines delivery formats, fidelity to source material, evidence requirements and user review before execution.
GlobalGPT ResearchPlatform activity statistics and AI-assisted qualitative analysis of selected request excerpts; findings do not estimate population prevalence.
Read the ResearchMove From Reading to Making.
Product entrances stay visually distinct from editorial guides, so every click has a clear destination.
01 / 06
01ChatGPT Image Generator
Generate new visuals directly from a focused image workspace.
GlobalGPT
02ChatGPT Photo Editor
Move directly into an image editing workflow.
GlobalGPT
03ChatGPT Caricature Generator
Turn a portrait idea into a more playful visual style.
GlobalGPT
04ChatGPT Rewriter
Refine tone and clarity from a focused workspace.
GlobalGPT
05ChatGPT Cover Letter Generator
Shape a focused cover letter around a role and experience.
GlobalGPT
06ChatGPT Checker
Review text in a dedicated checking workflow.
GlobalGPT
Know Which Voice You Are Reading.
Official model facts, public background and GlobalGPT editorial content stay visibly separate.
Official OpenAI
GPT-6 Astra Model DocumentationCapabilities · Checked Sep 14, 2026OpenAI API ChangelogOpenAI model and API updatesGPT Image 2.5 Sunburst DocumentationImage generation and editingGPT Image 2.5 Flare DocumentationEveryday image generation · Checked Sep 14, 2026GPT-6.1 Sol Model DocumentationComplex coding and professional workGPT-6 Luna Model DocumentationFocused, high-volume tasksOpenAI API PricingOfficial provider · 2026-10-28ChatGPT Work and Codex PlansOfficial provider · 2026-10-28GPT-6.1 Terra Model ReferenceOfficial provider · 2026-10-28xAI Grok 4.7 EvaluationOfficial provider · 2026-10-28Anthropic Haiku 5.5 EvaluationOfficial provider · 2026-10-28GPT-6.1 Terra: Official GPT Model Benchmark ComparisonsOpenAI · 2026-10-28ChatGPT Questions, Answered
Model choice, access routes and evidence in one place.
Which GPT model should I start with?
Start with Astra for demanding reasoning, Sol for coding and professional tasks, Terra for visual reasoning and writing, or Luna for focused work. These are workload recommendations based on official positioning.
Can GPT Image 2.5 generate and edit pictures?
Yes. The official models take text and image inputs and produce images. GlobalGPT offers one GPT Image 2.5 workspace; describe the requested creation or edit in the prompt.
Is a GlobalGPT subscription the same as ChatGPT Plus?
No. GlobalGPT provides its own model entries and account allowances. ChatGPT subscriptions buy OpenAI’s app features and usage under the chosen plan.
Are the API prices included in my subscription?
The API table quotes direct OpenAI developer billing in USD per million tokens. It does not set GlobalGPT prices or convert into included ChatGPT tasks.
Can these models browse or use a computer automatically?
Those actions require a workspace that exposes the relevant tools and permissions. A model’s documented capability does not prove that every chat entry includes those tools.
Are the benchmark results GlobalGPT tests?
No. These are OpenAI-published research/API evaluation results for four GPT models at Max reasoning effort. The table links to the source charts. Factual error rate uses deliberately difficult prompts and does not estimate everyday error rates; these results are not GlobalGPT measurements.