Dateline: March 26, 2026 | Next update: April 2, 2026
This guide covers four AI model families in equal depth. For each it explains not just the models but the full product ecosystem: tools, apps, integrations, pricing plans, and recent launches.
Claude (Anthropic)
Anthropic was founded in 2021 by former OpenAI researchers including Dario and Daniela Amodei. Their approach, called Constitutional AI, trains models to have internalised values rather than externally imposed rules. Claude is known for careful, nuanced responses and self-awareness about its limitations. As of March 2026 it holds the number-one position on the Chatbot Arena text leaderboard and leads on software engineering benchmarks. Claude Code, its coding agent, has surpassed $2.5 billion in annualised revenue as of March 2026, up from $1 billion in January.
Claude Haiku 4.5 — fast and cheap
Aug 2025 | $1/$5 per MTok | API: claude-haiku-4-5-20251001 | Haiku 3 retiring Apr 19 2026
The lightest and fastest Claude model. Best for high-volume tasks where speed matters more than depth: text classification, fact extraction, simple Q&A, and routing layers in automated pipelines that escalate harder queries to a stronger model.
Technical detailsContext: 200K tokens. SWE-bench: 73.3%. Batch API: 50% discount. Vision and tool use supported. Token consumption ~3–4x lower than Sonnet for comparable queries. Haiku 3 retires April 19 2026 — migrate to claude-haiku-4-5-20251001 before then.
Best for: High-volume pipelines, classification, simple Q&A
Claude Sonnet 4.6 — the everyday standard
Feb 17 2026 | $3/$15 per MTok | API: claude-sonnet-4-6 | Default free model on claude.ai
The model most people should use for most tasks. Became the default free model on claude.ai in February 2026. Handles complex analysis, long documents, structured writing, code generation, and multi-step reasoning. Adaptive reasoning automatically calibrates thinking depth to query complexity.
★ What's newBecame the default free model on claude.ai in February 2026. In user preference studies, 70% of Claude Code users preferred it to the previous generation; 59% preferred it over the prior Opus generation for coding.
Technical detailsContext: 200K standard / 1M beta. Output: up to 64K tokens. Adaptive extended thinking replaces manual thinking budget controls. Elo 1,633 (professional writing leaderboard). Prompt caching: ~10% of base input cost on cache hits. Available on AWS Bedrock and Google Vertex AI. Default model for all Cowork sessions.
Best for: Most professional use — writing, analysis, coding, long documents
Claude Opus 4.6 — the flagship
Feb 5 2026 | $5/$25 per MTok | API: claude-opus-4-6 | 67% price cut from prior generation ($15/$75)
Most powerful Claude model, for work demanding deep and sustained reasoning. Number one on Chatbot Arena (Elo 1504). Leads all published models on software engineering: 80.84% SWE-bench Verified. Task completion horizon: 14.5 hours of autonomous work (50th percentile), the longest published figure for any model. Introduced Agent Teams: native framework for coordinating multiple Claude instances simultaneously.
★ What's newAgent Teams: native multi-agent orchestration in a single API call. 1M token context window now GA as of March 2026 (unified pricing). Code Security (Enterprise/Team preview): data flow reasoning for vulnerability detection.
Technical detailsContext: 200K standard / 1M GA (March 2026). Max output: 128K tokens. SWE-bench Verified: 80.84%. METR task horizon: 14h30m (50th percentile). Used by Norway sovereign wealth fund for ESG screening (Feb 2026); NASA Mars rover routing (Dec 2025).
Best for: Deep reasoning, long-horizon autonomous tasks, complex agentic workflows
Claude Code — agentic coding tool
Platform: terminal / web / mobile / VS Code | Availability: Pro plan and above
A terminal-based coding agent giving Claude access to your codebase, terminal, and files. Writes code, runs tests, reads errors, and fixes bugs with minimal human input. Features include remote access, scheduled recurring tasks, a plugin marketplace (MCP integrations), parallel agents, and auto memory (persistent project architecture knowledge across sessions).
★ What's new (March 2026)Auto mode (Mar 24): AI safety classifier approves routine developer actions automatically and screens for prompt injection attacks. Claude Code Review: automatic background code reviewer to catch bugs before they reach the codebase. Claude Code Channels (Mar 24): control Claude Code via Discord and Telegram.
Technical detailsSWE-bench terminal agent: ~75%. Remote access: browser/mobile session control. Plugin marketplace: MCP integrations with Team/Enterprise admin controls. Auto mode: safety classifier with prompt injection screening; small token/latency overhead. Revenue: $2.5B+ annualised March 2026, up from $1B in January.
Best for: Software engineering with full agentic control
Claude Cowork — for non-developers
Launched: Jan 2026 (Mac), Feb 2026 (Windows) | Availability: Pro, Team, Enterprise
Gives Claude persistent access to your local files in a sandboxed virtual machine. Connects to Gmail, Google Drive, Google Calendar, Slack, and 38+ services. Projects create isolated workspaces per context with their own files and history. Cross-session memory learns your preferences over time.
★ What's new (March 2026)Computer use (Mar 23): Claude can now control your Mac — clicking, typing, scrolling, opening apps, navigating browsers, filling forms. Connector-first approach uses structured API integrations when available; falls back to keyboard-and-mouse only when no connector exists. Financial/trading/crypto apps blocked by default. macOS research preview; Pro and Max subscribers only.
★ What's new (March 2026)Dispatch (Mar 17): Persistent conversation thread across phone and Claude Desktop. Assign tasks from phone; Claude executes on your desktop in the background. Scheduled recurring and on-demand tasks: set tasks to run automatically — weekly summaries, daily briefings, recurring reports — without prompting each time.
Technical detailsSandboxed VM for file access; computer use runs OUTSIDE the VM on the actual macOS desktop. Per-app permission model; blocklist user-configurable. 38+ native MCP connectors. Projects: isolated context per folder. Memory excludes passwords, health, financial data by default. Requires Pro ($20/mo), Max ($100–200/mo), Team, or Enterprise. Computer use: macOS only, research preview. Sonnet 4.6 is default; user can switch to Opus 4.6.
Best for: Knowledge workers who want AI running multi-step tasks across files and apps
Claude in Excel and PowerPoint
Platform: Microsoft Office | Availability: Team / Enterprise | Supports: Amazon Bedrock, Google Vertex AI, Microsoft Foundry via LLM gateway
Add-in for Microsoft Excel (Opus 4.6, native pivot table editing and conditional formatting) and add-in for PowerPoint (creates, edits, and revises slides). Both share the full open document as context and support custom skills and connectors via LLM gateway integrations with major cloud providers.
Best for: Microsoft Office users, enterprise knowledge workers
Health data on mobile
Claude can read and analyse health and fitness data on iOS and Android (activity patterns, workouts, sleep). Available on Pro and Max plans, currently US-only. Android requires Health Connect and Android 14+.
Plans and Pricing
| Plan |
Price |
Key Includes |
| Free |
$0 |
Sonnet 4.6, web search, memory, Artifacts, code execution, image analysis |
| Pro |
$20/month |
5x usage, full Claude Code, unlimited Projects, Research mode, Cowork (incl. computer use preview) |
| Max |
$100–200/month |
All Pro features, higher limits, early feature access, priority at peak |
| Team |
$25/seat/month (annual) |
Min 5 seats. Admin controls, shared Projects, SSO, MS365 + Slack |
| Enterprise |
Custom |
500K context, HIPAA readiness, SCIM, audit logs. Self-serve purchase available |
Technical details200K rule: input above 200K tokens billed at 2x ($3→$6 input, $15→$22.50 output per MTok). Batch API: 50% discount. Prompt caching: up to 90% off repeated context. Developer math: $200 Max is ~18x cheaper than equivalent direct API token spend for heavy Claude Code use. Claude 5 (codename Fennec) has appeared in Google Vertex AI infra logs — expected mid-2026.
ChatGPT / OpenAI
ChatGPT is OpenAI's general-purpose AI assistant. It is a single place where people ask questions, draft and edit writing, analyse files, search the web, generate images and video, write code, talk by voice, and delegate multi-step work to an agent. The product routes between several models and tools depending on what you ask it to do.
GPT-5.3 Instant — the default everyday model
Released: 3 March 2026 | Availability: all ChatGPT tiers
The fast general-purpose default model for chat, everyday writing, search, and tool use. Supports the widest set of ChatGPT tools and is the baseline experience on Free as well as paid plans.
★ What's new16 March 2026: update improves follow-up tone and reduces teaser-style phrasing.
Technical detailsRate limits: Free — 10 messages per 5 hours, then fallback to mini. Plus and Go — 160 messages per 3 hours. Business and Pro described as effectively unlimited. Context: 16K (Free), 32K (Plus/Business), 128K (Pro/Enterprise). Tools: web search, data analysis, image analysis, file analysis, canvas, image generation, memory, custom instructions.
Best for: General chat, fast drafting, routine search, file Q&A, broadest day-to-day tool-compatible use
GPT-5.4 Thinking — the flagship reasoning model
Released: 5 March 2026 | API: $2.50/$15 per 1M tokens | Availability: Plus, Business, Pro, Enterprise, Edu
OpenAI's current flagship reasoning model. Slower and more expensive than the default Instant model, but better at multi-step reasoning, coding, document-heavy work, and difficult research questions. The model to pick when quality matters more than speed.
★ What's newLaunched 5 March 2026, replacing GPT-5.2 Thinking as the current flagship reasoning path.
Technical detailsAPI ID: gpt-5.4. Context: 1M tokens. Max output: 128K tokens. API pricing: Standard $2.50 input / $0.25 cached / $15 output per 1M tokens. Flex: $1.25 / $0.13 / $7.50. Batch: half standard. Priority: $5 / $0.50 / $30. Benchmarks: OSWorld 75.0; SWE-bench Pro 57.7; GPQA Diamond 92.8; ARC-AGI-2 73.3; Harvey BigLaw Bench 91%; Humanity's Last Exam 52.1 with tools.
Best for: Complex writing, multi-step analysis, coding, legal and research work
GPT-5.4 Pro — the highest-quality option
Released: 5 March 2026 | API: $30/$180 per 1M tokens | Availability: ChatGPT Pro, Business/Enterprise/Edu
The version for users willing to spend far more for a better answer on especially difficult tasks. Not a general default — the 'use this when the stakes are high and the problem is hard' option. Note: Apps, Memory, Canvas, and image generation are not available with the Pro model in ChatGPT.
★ What's newLaunched 5 March 2026 alongside GPT-5.4 Thinking, with higher benchmark scores on several hard reasoning tests but a narrower ChatGPT tool set.
Technical detailsAPI ID: gpt-5.4-pro. Context: 1,050,000 tokens. Max output: 128K tokens. API pricing: $30 input / $180 output per 1M tokens standard. For prompts above 272K tokens: 2x input and 1.5x output. Regional endpoints: 10% uplift. Benchmarks: GPQA Diamond 94.4; ARC-AGI-2 83.3; Humanity's Last Exam 58.7 with tools; BrowseComp 89.3. Some API requests can take minutes — background mode recommended.
Best for: Very hard reasoning problems, high-stakes review, cases where cost matters less than quality
GPT-5.4 mini — the smaller reasoning model
Released: 17 March (API/Codex), 18 March (ChatGPT) | API: $0.75/$4.50 per 1M tokens
A cheaper API model and practical fallback in the ChatGPT user experience. For Free and Go users it appears through the Thinking feature. For most paid users it is the model ChatGPT falls back to when GPT-5.4 Thinking limits are reached.
★ What's newLaunched 17–18 March 2026 as both a user-facing Thinking option for Free/Go and a fallback model for paid users hitting GPT-5.4 Thinking limits.
Technical detailsAPI ID: gpt-5.4-mini. Context: 400K tokens. API pricing: $0.75 input / $0.075 cached / $4.50 output per 1M tokens. Batch: half standard. Benchmarks: SWE-bench Pro 54.4; Terminal-Bench 60.0; MCP Atlas 57.7.
Best for: Budget-sensitive reasoning, graceful fallback, high-volume API workloads needing a strong reasoning model
GPT-5.4 nano — the API-only lowest-cost model
Released: 17 March 2026 | API: $0.20/$1.25 per 1M tokens | API only
Not a mainstream ChatGPT model. Exists for developers who need a very cheap, very fast model for automation, classification, extraction, and high-volume backend tasks.
★ What's newLaunched 17 March 2026 as the cheapest current GPT-5.4-class API model.
Technical detailsAPI ID: gpt-5.4-nano. Context: 400K tokens. API pricing: $0.20 input / $0.02 cached / $1.25 output per 1M tokens. Benchmarks: SWE-bench Pro 52.4; Terminal-Bench 46.3; MCP Atlas 56.1.
Best for: Developer automation, classification, extraction, cost-sensitive inference pipelines
Legacy models — what remains and until when
GPT-5.2 Thinking remains in Legacy Models for Plus and Pro users for 90 days after the 5 March 2026 launch — retiring 5 June 2026. GPT-4o in Custom GPTs: Business, Enterprise, and Edu retain it until 3 April 2026. Already retired from ChatGPT (13 Feb 2026): GPT-4o, GPT-4.1, GPT-4.1 mini, OpenAI o4-mini, and GPT-5 (Instant and Thinking).
Agent Mode — ChatGPT that can act
Broad rollout live by March 2026 | Included on Plus, Pro, Business, Enterprise, Edu
Instead of just replying with advice, Agent Mode can browse websites, work through forms, inspect files, use tools, and carry out multi-step tasks. Operates through OpenAI's agent environment. When a site needs a login or sensitive confirmation, control is handed back to the user. Typical task duration: 5–30 minutes.
Technical detailsLimits: Plus 40 agent messages/month; Pro 400/month; Business/Enterprise 40/month. Business flexible pricing: 30 credits per message. User remains in control for logins — screenshots not captured while user controls browser.
Best for: Delegating long web tasks, repetitive information gathering, form-heavy work
Canvas — shared editing workspace
Availability: web, Windows, macOS; iOS/Android coming soon | Not available with GPT-5.4 Pro
A persistent canvas where text or code can be revised, marked up, and iterated on directly. Supports Python code execution, can be enabled inside GPTs, and allows sharing of canvas assets. Most useful when a task has become too large for ordinary chat bubbles.
Best for: Long drafts, iterative editing, structured code review
Sora 2 — video generation
Launched: 30 September 2025 | Available via sora.com, iOS, Android, and ChatGPT plans
OpenAI's current video generation system. Consumer limits: Plus/Business — unlimited images and video, up to 480p and 10 seconds, 1 concurrent generation. Pro — unlimited, faster, up to 1080p and 20 seconds, up to 5 concurrent, watermark-free.
★ What's new12 March 2026: Sora API expanded with reusable character references, longer 20-second generations, 1080p output for sora-2-pro, video extensions, and Batch API support.
Technical detailsAPI IDs: sora-2 and sora-2-pro. API pricing: sora-2 at 720p is $0.10/sec standard. sora-2-pro: $0.30/sec at 720p, $0.50/sec at 1024p, $0.70/sec at 1080p; batch is half standard prices.
Best for: Creative video generation, concept visualization, marketing drafts, storyboarding
Deep Research — long-form, citation-backed research
Available in ChatGPT | Outputs: Markdown, Word, or PDF
Creates a research plan, works through sources over time, and returns a structured report with citations. Connected apps are read-only. Users can edit the research plan before the run starts. Legacy Deep Research mode removed 26 March 2026.
★ What's new19 March 2026: OpenAI added editable research plan and improved report view. Legacy Deep Research mode removed 26 March 2026.
Best for: Policy briefs, market research, literature-style reviews, multi-source synthesis
Memory, Projects, and Library — the persistence layer
Memory and Projects live | Library launched 23 March 2026
Memory lets ChatGPT remember past details. Projects group chats and files together. Library is the new filing cabinet for uploaded and generated files — reusable across chats.
★ What's new23 March 2026: Library launched for saved files, making uploaded and created files reusable across chats. Currently web-only; recent files in composer and file search on iOS and Android. Available to Plus, Pro, and Business users outside the EEA, Switzerland, and UK.
Best for: Ongoing projects, recurring work, personalized assistance
OpenAI Codex — the coding platform
App launched: 2 February 2026 | Included in Plus, Pro, Business, Edu, Enterprise
A dedicated coding environment with multiple parallel agents, isolated worktrees, reviewable diffs, and delegation to cloud agents. Integrations: GitHub, Slack, Linear.
★ What's new4 March 2026: Expanded on Windows for Business workspaces. 10 March: auto top-up for shared credits used by Codex and Sora. GPT-5.3-Codex remains the specialized coding model launched on 5 February 2026.
Best for: Serious software work, parallel coding tasks, teams wanting agentic help in real development workflows
Plans and Pricing
| Plan |
Price |
Key Includes |
| Free |
$0 |
Limited GPT-5.3, limited uploads and image generation, limited Deep Research |
| Go |
$8/month (US) |
More GPT-5.3, more uploads and images, longer memory. May include ads. |
| Plus |
$20/month |
Advanced reasoning models, projects, tasks, Codex, Sora, expanded Deep Research and Agent Mode |
| Pro |
$200/month |
GPT-5.4 Pro, maximum Deep Research and Agent Mode, expanded Sora, higher-priority Codex |
| Business (formerly Team) |
$25/user/month annual |
Shared workspace, admin controls, Codex, agent access. Min 2 users. |
| Enterprise |
Custom |
Bigger context/file support, SCIM, RBAC, EKM, data residency, SLAs |
Technical detailsAPI pricing (current GPT-5 family): gpt-5.4 $2.50/$0.25 cached/$15 output per 1M tokens; gpt-5.4-pro $30/$180; gpt-5.4-mini $0.75/$0.075/$4.50; gpt-5.4-nano $0.20/$0.02/$1.25. API billing is separate from ChatGPT subscriptions. Self-serve API usage tiers scale from ~$100/month at low tiers to $200,000/month at Tier 5.
Gemini (Google)
Gemini is Google's primary artificial intelligence system — a single "brain" that can see, hear, read, and create. It is multimodal: it understands text, images, audio, and video at once. It powers everything from a simple chat box on your phone to professional tools used by scientists and developers.
Gemini 3.1 Flash Lite — the efficiency specialist
March 3, 2026 | $0.25 / 1M tokens | Developers & Enterprise
The newest and most affordable member of the family. Built for high-volume work where you need the AI to do a simple task thousands of times per second without costing a fortune: sorting, translating short messages, labeling photos.
★ What's newLaunched March 3, 2026, as the first "Lite" model in the 3.1 generation, optimized for latency-sensitive production pipelines. 2.5x faster Time to First Token (TTFT) than Gemini 2.5 Flash.
Technical detailsContext Window: 1M tokens (32K via API Preview). API ID: gemini-3.1-flash-lite-preview. API Pricing: $0.25 per 1M input / $1.50 per 1M output. Benchmarks: GPQA Diamond: 86.9% | MMMU-Pro: 76.8% | MMMLU: 88.9%.
Best for: High-volume data processing, latency-sensitive production pipelines
Gemini 3.1 Flash — the balanced workhorse
February 2026 | $0.50 / 1M tokens | All users (Free & Paid)
The default version of Gemini for most people. Strikes the best balance between being clever and being quick. Handles vacation planning, PDF summaries, email drafting, and real-time conversation.
Technical detailsContext Window: 1M tokens. API Pricing: $0.50 per 1M input / $3.00 per 1M output. API ID: gemini-3.1-flash. Benchmarks: ARC-AGI-2: 45.2% | MMLU-Pro: 82.1%.
Best for: General-purpose chat, speed-sensitive everyday tasks
Gemini 3.1 Pro — the flagship
February 19, 2026 | $2.00 / 1M tokens | AI Pro & Ultra subscribers
Google's most capable model for reasoning. Designed for hard problems: debugging complex code, synthesizing conflicting data, creating intricate live dashboards or animated graphics. Features "Extended Thinking" — pauses to reason before answering difficult logic puzzles.
Technical detailsContext Window: 1M to 2M tokens. Output Window: 65,536 tokens. API Pricing: $2.00 per 1M input / $12.00 per 1M output (up to 200k context); rates double above 200k. Benchmarks: ARC-AGI-2: 77.1% (SOTA) | GPQA Diamond: 94.3% | Terminal-Bench 2.0: 68.5%.
Best for: Complex logic and coding, scientific research, multi-source synthesis
Gemini 3.1 Ultra — the powerhouse
February 2026 | Subscription Only ($249.99/mo) | AI Ultra subscribers
The top-tier model, exclusive to the most expensive plan. Used for the most demanding creative tasks: generating professional 4K video with Veo 3.1, running advanced Deep Research reports that take up to 20 minutes to compile.
Technical detailsCurrently holds #1 on Artificial Analysis Intelligence Index (Score: 61). API access restricted — primarily via Vertex AI for enterprise and through the Gemini App for Ultra subscribers.
Best for: State-of-the-art research, 4K video generation, highest-stakes creative work
NotebookLM — personalized research assistant
Upload your own documents (PDFs, transcripts, websites), and NotebookLM creates a private world where the AI only answers based on your specific information. Audio Overviews: two AI "hosts" discuss your documents like a podcast. Study Suite: quizzes, flashcards, and study guides from your sources.
★ What's newCinematic Video Overviews (March 15, 2026): launched for Ultra tier users to create 4K visual summaries. Mind Maps and Artifacts: automatically generates visual diagrams of how your ideas connect.
Technical detailsSource Limits: up to 50 sources per notebook; 500k words per source. Export: supports .pptx and .docx artifact creation.
Best for: Research and study, private knowledge-base Q&A
Jules — the AI coding agent
An asynchronous coding partner. You give Jules a big task, close your laptop, and Jules works in the background — cloning your code, testing its own work, and sending you a pull request when it's finished.
Technical detailsBase Model: Gemini 2.5 Pro / 3.1 Pro. Limits: Free (15 tasks/day), Pro (100 tasks/day), Ultra (300 tasks/day).
Best for: Autonomous coding, background task execution
Veo 3.1 and Lyria 3
- Veo 3.1: Google's 4K video generation model. Creates 8-second cinematic clips with synchronized audio.
- Lyria 3: High-fidelity music generation. Creates 3-minute songs with vocals and lyrics from a text prompt.
★ What's newLyria 3 (March 25, 2026): updated to support 3-minute professional arrangements with realistic vocals.
Plans and Pricing
| Plan |
Price |
Key Includes |
Not Included |
| Basic (Free) |
$0.00 |
Gemini 3.1 Flash, 32k Context |
Deep Research, Jules, 4K Video |
| AI Plus |
$7.99/mo |
200GB Storage, 200 Credits |
3.1 Pro Model, High Jules limits |
| AI Pro |
$19.99/mo |
2TB Storage, Gemini 3.1 Pro |
Veo 3.1 Standard (4K) |
| AI Ultra |
$249.99/mo |
30TB Storage, Gemini 3.1 Ultra |
— |
Microsoft Copilot
Microsoft Copilot is not a single AI model. It is a family of AI-powered features built on top of models from other companies (primarily OpenAI GPT-5 and Anthropic Claude). What makes Copilot distinctive is its Work IQ, its integration with Microsoft Graph, and its ability to live inside Office apps like Word, Excel, and Teams. The underlying models provide raw reasoning and language capabilities, while Copilot adds organizational context, compliance, and productivity integration.
Copilot Free
Pricing: Free
Basic chat and writing assistance inside Microsoft Edge and Office web apps. Limited to general-purpose queries and does not access organizational data.
Technical detailsUses GPT-4.5. No Microsoft Graph integration. No compliance frameworks. No admin controls.
Best for: Individual users wanting lightweight AI help without enterprise integration
Copilot Pro
Pricing: $20/month
Unlocks advanced writing, summarization, and design features in Word, Excel, PowerPoint, and Outlook. Requires a separate Microsoft 365 subscription for full app integration.
Technical detailsAccess to GPT-5 and Claude 3. Limited Graph integration. No enterprise compliance. Admin controls not available.
Best for: Power users and freelancers who need deeper AI integration in Office apps
Copilot Business
Pricing: $30/user/month
Integrates with Microsoft Graph, enabling context-aware assistance across Teams, SharePoint, and OneDrive. Includes meeting transcription, AI notes, and live translation.
Technical detailsGraph integration: SharePoint, OneDrive, Teams. Compliance: ISO 27001, SOC 2. Admin portal with role-based controls.
Best for: Small and medium businesses needing collaboration and compliance features
Copilot Enterprise
Pricing: $30/user/month (requires E3/E5 licence)
Adds Semantic Index, Purview governance, compliance APIs, SCIM provisioning, and vertical add-ons like Copilot for Sales and Service.
Technical detailsFull Graph integration. Semantic Index for organizational knowledge. Purview compliance. SCIM support. API access for enterprise developers.
Best for: Large organizations requiring compliance, governance, and extensibility
E7 Tier
Pricing: $57/user/month
Bundles Microsoft 365 E7 with Copilot Enterprise. Includes advanced security, compliance, and analytics.
Best for: Enterprises standardizing on Microsoft's highest security and compliance tier
Agent 365
Pricing: $15/user/month
A lightweight agent framework for workflow automation. Enables Copilot to act across apps and services without full Microsoft 365 integration. Claude 3-based agent.
Best for: Teams needing automation without full enterprise licensing
Copilot Cowork
Pricing: Research preview
A collaborative AI agent built with Anthropic. Emphasizes Work IQ — the ability to track tasks, checkpoints, and plan-to-action loops.
★ What's newMarch 18, 2026: Cowork added inbox and calendar awareness in Copilot Chat.
Technical detailsClaude Cowork architecture. Graph integration optional. Compliance: SOC 2, ISO 27001. Enterprise Cowork adds governance APIs.
Best for: Teams experimenting with collaborative AI workflows
GitHub Copilot
Pricing: Free (students/OSS), Pro $10/month, Pro+ $15/month, Business $19/user/month, Enterprise $39/user/month
Assists with code completion, multi-file reasoning, and documentation. Enterprise plans add custom knowledge bases and fine-tuned models.
Technical detailsMulti-file agent. Enterprise supports private model fine-tuning. Compliance: SOC 2, ISO 27001.
Best for: Developers and enterprises needing AI-powered coding assistance
Copilot inside Microsoft 365 Apps
- Word: Drafting, summarization, rewriting
- Excel: Data analysis, formula generation, chart creation
- PowerPoint: Slide generation, design suggestions
- Outlook: Email drafting, summarization
- Teams: Meeting transcription, AI notes, live translation
- SharePoint: Content summarization, metadata tagging
- OneDrive: File summarization, search assistance
★ What's new (March 2026)Excel: predictive trend analysis added.
Best for: Microsoft 365 users wanting AI embedded directly in productivity apps
Plans and Pricing
| Tier |
Price |
Requirements |
Models |
| Copilot Free |
$0 |
Microsoft account |
GPT-4.5 |
| Copilot Pro |
$20/month |
Microsoft 365 subscription |
GPT-5, Claude 3 |
| Copilot Business |
$30/user/month |
5 seats minimum |
GPT-5, Claude 3 |
| Copilot Enterprise |
$30/user/month |
E3/E5 licence |
GPT-5, Claude 3 |
| E7 Tier |
$57/user/month |
Enterprise licence |
GPT-5 |
| Agent 365 |
$15/user/month |
None |
Claude 3 |
| GitHub Copilot Pro |
$10/month |
GitHub account |
GPT-4.5 |
| GitHub Copilot Enterprise |
$39/user/month |
Enterprise licence |
GPT-5 |
This document is kept current as of March 2026. For the most up-to-date figures, please request a re-run of this guide monthly.
Datum: 26. mart 2026. | Sledeće izdanje: 2. april 2026.
Ovaj vodič pokriva četiri porodice AI modela u jednakoj dubini. Za svaku objašnjava ne samo modele, već i kompletan proizvodni ekosistem: alate, aplikacije, integracije, planove i cene i nedavna lansiranja.
Claude (Anthropic)
Anthropic je osnovan 2021. od strane bivših OpenAI istraživača, uključujući Daria i Danielu Amodei. Njihov pristup, poznat kao Constitutional AI, trenira modele da imaju internalizovane vrednosti umesto spolja nametnutih pravila. Claude je poznat po pažljivim, nijansiranim odgovorima i svesnosti o sopstvenim ograničenjima. Od marta 2026. drži prvo mesto na Chatbot Arena liderbordu za tekst i vodi na benčmarkovima softverskog inženjeringa. Claude Code, njegov agent za kodiranje, premašio je 2,5 milijardi dolara godišnjeg prihoda od marta 2026, u porastu sa milijarde dolara u januaru.
Claude Haiku 4.5 — brz i jeftin
Avg. 2025. | $1/$5 po MTok | API: claude-haiku-4-5-20251001 | Haiku 3 se povlači 19. apr. 2026.
Najlakši i najbrži Claude model. Najpogodniji za zadatke velikog obima gde je brzina važnija od dubine: klasifikacija teksta, ekstrakcija činjenica, jednostavni Q&A i slojevi rutiranja u automatizovanim procesima koji eskaliraju teže upite ka snažnijem modelu.
Tehnički detaljiKontekst: 200K tokena. SWE-bench: 73,3%. Batch API: 50% popust. Podržava vid i upotrebu alata. Potrošnja tokena ~3–4x niža od Sonnet-a za uporedive upite. Haiku 3 se povlači 19. aprila 2026. — migrirajte na claude-haiku-4-5-20251001 pre tog datuma.
Najpogodnije za: Procesi velikog obima, klasifikacija, jednostavni Q&A
Claude Sonnet 4.6 — svakodnevni standard
17. feb. 2026. | $3/$15 po MTok | API: claude-sonnet-4-6 | Podrazumevani besplatni model na claude.ai
Model koji bi većina ljudi trebalo da koristi za većinu zadataka. Od februara 2026. je podrazumevani besplatni model na claude.ai. Rukuje složenom analizom, dugim dokumentima, strukturiranim pisanjem, generisanjem koda i višekoračnim zaključivanjem. Adaptivno rezonovanje automatski prilagođava dubinu razmišljanja složenosti upita.
★ Šta je novoPostao podrazumevani besplatni model na claude.ai u februaru 2026. U korisničkim istraživanjima, 70% korisnika Claude Code-a preferiralo ga je u odnosu na prethodnu generaciju; 59% ga je preferiralo u odnosu na prethodnu Opus generaciju za kodiranje.
Tehnički detaljiKontekst: 200K standardno / 1M beta. Izlaz: do 64K tokena. Adaptivno prošireno razmišljanje zamenjuje ručne kontrole budžeta razmišljanja. Elo 1.633 (liderboard za profesionalno pisanje). Keš upita: ~10% od bazične ulazne cene na pogocima keša. Dostupno na AWS Bedrock i Google Vertex AI. Podrazumevani model za sve Cowork sesije.
Najpogodnije za: Većina profesionalne upotrebe — pisanje, analiza, kodiranje, dugi dokumenti
Claude Opus 4.6 — flagship model
5. feb. 2026. | $5/$25 po MTok | API: claude-opus-4-6 | Sniženje cene za 67% u odnosu na prethodnu generaciju ($15/$75)
Najmoćniji Claude model, za zadatke koji zahtevaju duboko i dugotrajno rezonovanje. Broj jedan na Chatbot Arena (Elo 1504). Vodi sve objavljene modele u softverskom inženjeringu: 80,84% SWE-bench Verified. Horizont dovršavanja zadataka: 14,5 sati autonomnog rada (50. percentil), najduža objavljena cifra za bilo koji model. Uveo Agent Teams: nativni okvir za koordinaciju više Claude instanci istovremeno.
★ Šta je novoAgent Teams: nativna orchestracija više agenata u jednom API pozivu. Kontekstni prozor od 1M tokena sada je GA od marta 2026. (ujednačeno određivanje cena). Code Security (Enterprise/Team pregled): analiza toka podataka za otkrivanje ranjivosti.
Tehnički detaljiKontekst: 200K standardno / 1M GA (mart 2026.). Maksimalni izlaz: 128K tokena. SWE-bench Verified: 80,84%. METR horizont zadataka: 14h30min (50. percentil). Korišćen od strane norveškog suverenog fonda za ESG pregled (feb. 2026.); NASA rutiranje marsovskog rovera (dec. 2025.).
Najpogodnije za: Duboko rezonovanje, dugoročni autonomni zadaci, složeni agentski procesi
Claude Code — agent za kodiranje
Platforma: terminal / veb / mobilni / VS Code | Dostupnost: Pro plan i više
Agent za kodiranje baziran na terminalu koji Claude-u daje pristup bazi koda, terminalu i fajlovima. Piše kod, pokreće testove, čita greške i ispravlja bagove uz minimalan ljudski doprinos. Funkcionalnosti: daljinski pristup, zakazani ponavljajući zadaci, marketplace dodataka (MCP integracije), paralelni agenti i automatska memorija (trajno znanje o arhitekturi projekta između sesija).
★ Šta je novo (mart 2026.)Auto mod (24. mar.): AI klasifikator bezbednosti automatski odobrava rutinske akcije programera i filtrira napade ubacivanjem upita. Claude Code Review: automatski pregledač koda u pozadini koji hvata bagove pre nego što dospeju u bazu koda. Claude Code Channels (24. mar.): upravljanje Claude Code-om putem Discord-a i Telegram-a.
Tehnički detaljiSWE-bench terminal agent: ~75%. Daljinski pristup: kontrola sesije iz pregledača/mobilnog. Marketplace dodataka: MCP integracije sa Team/Enterprise admin kontrolama. Auto mod: klasifikator bezbednosti sa filtriranjem ubacivanja upita; mali overhead u tokenima/kašnjenju. Prihod: 2,5B+ godišnje u martu 2026., u porastu sa 1B u januaru.
Najpogodnije za: Softverski inženjering sa punom agentskom kontrolom
Claude Cowork — za nekodere
Lansiran: jan. 2026. (Mac), feb. 2026. (Windows) | Dostupnost: Pro, Team, Enterprise
Daje Claude-u trajan pristup vašim lokalnim fajlovima u izolovanoj virtuelnoj mašini. Povezuje se sa Gmail-om, Google Drive-om, Google kalendarom, Slack-om i 38+ servisa. Projekti kreiraju izolovane radne prostore po kontekstu sa sopstvenim fajlovima i istorijom. Memorija kroz sesije uči vaše preferencije tokom vremena.
★ Šta je novo (mart 2026.)Korišćenje računara (23. mar.): Claude sada može upravljati vašim Mac-om — klik, kucanje, skrolovanje, otvaranje aplikacija, navigacija pregledačima, popunjavanje formulara. Pristup koji daje prioritet konektorima koristi strukturirane API integracije kada su dostupne; prelazi na tastaturu i miš samo kada ne postoji konektor. Finansijske/trading/kripto aplikacije blokirane su po podrazumevanom podešavanju. Istraživački pregled za macOS; samo za Pro i Max pretplatnike.
★ Šta je novo (mart 2026.)Dispatch (17. mar.): Trajna konverzaciona nit između telefona i Claude Desktop-a. Dodeljujte zadatke sa telefona; Claude ih izvršava na vašem desktopu u pozadini. Zakazani ponavljajući i zadaci na zahtev: postavite zadatke da se automatski izvršavaju — nedeljni sažeci, dnevni brifovi, ponavljajući izveštaji — bez ponovnog pokretanja.
Tehnički detaljiIzolovana VM za pristup fajlovima; korišćenje računara se odvija VAN VM-a na stvarnom macOS desktopu. Model dozvola po aplikaciji; crna lista se može prilagoditi. 38+ nativnih MCP konektora. Projekti: izolovan kontekst po folderu. Memorija ne uključuje lozinke, zdravstvene i finansijske podatke po podrazumevanom podešavanju. Zahteva Pro ($20/mes.), Max ($100–200/mes.), Team ili Enterprise. Korišćenje računara: samo macOS, istraživački pregled. Sonnet 4.6 je podrazumevan; korisnik može preći na Opus 4.6.
Najpogodnije za: Radnici znanja koji žele AI koji obavlja višekoračne zadatke kroz fajlove i aplikacije
Claude u Excel-u i PowerPoint-u
Platforma: Microsoft Office | Dostupnost: Team / Enterprise | Podržava: Amazon Bedrock, Google Vertex AI, Microsoft Foundry
Dodatak za Microsoft Excel (Opus 4.6, nativno uređivanje pivot tabela i uslovnog formatiranja) i dodatak za PowerPoint (kreira, uređuje i revidira slajdove). Oba dele kompletan otvoreni dokument kao kontekst i podržavaju prilagođene veštine i konektore.
Najpogodnije za: Korisnici Microsoft Office-a, radnici znanja u preduzećima
Zdravstveni podaci na mobilnom
Claude može čitati i analizirati zdravstvene i fitnes podatke na iOS-u i Android-u (obrasci aktivnosti, treninzi, san). Dostupno na Pro i Max planovima, trenutno samo u SAD-u. Android zahteva Health Connect i Android 14+.
Planovi i cene
| Plan |
Cena |
Šta uključuje |
| Free |
$0 |
Sonnet 4.6, pretraga weba, memorija, Artifacts, izvršavanje koda, analiza slika |
| Pro |
$20/mes. |
5x korišćenje, pun Claude Code, neograničeni Projekti, Research mod, Cowork (uklj. pregled korišćenja računara) |
| Max |
$100–200/mes. |
Sve Pro funkcionalnosti, viši limiti, rani pristup funkcijama, prioritet u vršnim satima |
| Team |
$25/sedištu/mes. (godišnje) |
Min. 5 sedišta. Admin kontrole, deljeni Projekti, SSO, MS365 + Slack |
| Enterprise |
Po dogovoru |
500K kontekst, HIPAA pripremljenost, SCIM, revizijski zapisi. Kupovina bez prodajnog poziva. |
Tehnički detaljiPravilo 200K: ulaz iznad 200K tokena naplaćuje se po 2x ceni ($3→$6 ulaz, $15→$22,50 izlaz po MTok-u). Batch API: 50% popust. Keš upita: do 90% popusta za ponavljani kontekst. Za developere: $200 Max je ~18x jeftiniji od ekvivalentnog direktnog trošenja tokena API-ja za intenzivnu upotrebu Claude Code-a. Claude 5 (kodni naziv Fennec) pojavio se u Google Vertex AI infrastrukturnim logovima — očekuje se sredinom 2026.
ChatGPT / OpenAI
ChatGPT je OpenAI-jev asistent za opštu namenu. To je jedno mesto gde ljudi postavljaju pitanja, pišu i uređuju tekstove, analiziraju fajlove, pretražuju veb, generišu slike i video, pišu kod, razgovaraju glasom i sve više delegiraju višekoračni rad agentu. Proizvod rutira između nekoliko modela i alata u zavisnosti od toga šta tražite.
GPT-5.3 Instant — podrazumevani svakodnevni model
Objavljen: 3. mart 2026. | Dostupnost: svi ChatGPT nivoi
Brzi model opšte namene za razgovor, svakodnevno pisanje, pretragu i upotrebu alata. Podržava najširi skup ChatGPT alata i osnovno je iskustvo na besplatnom kao i na plaćenim planovima.
★ Šta je novo16. mart 2026.: ažuriranje poboljšava ton nastavka razgovora i smanjuje formulaičke fraze.
Tehnički detaljiOgraničenja: Free — 10 poruka na 5 sati, zatim prelaz na mini. Plus i Go — 160 poruka na 3 sata. Business i Pro opisani kao efektivno neograničeni. Kontekst: 16K (Free), 32K (Plus/Business), 128K (Pro/Enterprise). Alati: pretraga weba, analiza podataka, analiza slika, analiza fajlova, canvas, generisanje slika, memorija, prilagođena uputstva.
Najpogodnije za: Opšti razgovor, brzo pisanje, rutinska pretraga, Q&A nad fajlovima
GPT-5.4 Thinking — flagship model za rezonovanje
Objavljen: 5. mart 2026. | API: $2,50/$15 po 1M tokena | Dostupnost: Plus, Business, Pro, Enterprise, Edu
Trenutni OpenAI-jev flagship model za rezonovanje. Sporiji i skuplji od podrazumevanog Instant modela, ali bolji u višekoračnom zaključivanju, kodiranju, radu s dokumentima i teškim istraživačkim pitanjima.
★ Šta je novoLansiran 5. marta 2026., zamenjuje GPT-5.2 Thinking kao trenutni flagship put rezonovanja.
Tehnički detaljiAPI ID: gpt-5.4. Kontekst: 1M tokena. Maks. izlaz: 128K tokena. API cena: Standardno $2,50 ulaz / $0,25 keš / $15 izlaz po 1M tokena. Flex: $1,25 / $0,13 / $7,50. Batch: pola standardne. Prioritet: $5 / $0,50 / $30. Benčmarkovi: OSWorld 75,0; SWE-bench Pro 57,7; GPQA Diamond 92,8; ARC-AGI-2 73,3; Harvey BigLaw Bench 91%; Humanity's Last Exam 52,1 sa alatima.
Najpogodnije za: Složeno pisanje, višekoračna analiza, kodiranje, pravni i istraživački rad
GPT-5.4 Pro — opcija najvišeg kvaliteta
Objavljen: 5. mart 2026. | API: $30/$180 po 1M tokena | Dostupnost: ChatGPT Pro, Business/Enterprise/Edu
Verzija za korisnike koji su spremni da potroše mnogo više za bolji odgovor na posebno teške zadatke. Nije opšti podrazumevani model — to je opcija 'koristite ovo kada su ulog visoki i problem težak'. Napomena: Apps, Memory, Canvas i generisanje slika nisu dostupni sa Pro modelom u ChatGPT-u.
★ Šta je novoLansiran 5. marta 2026. zajedno sa GPT-5.4 Thinking, sa višim benčmark rezultatima na nekoliko teških testova rezonovanja, ali užim skupom ChatGPT alata.
Tehnički detaljiAPI ID: gpt-5.4-pro. Kontekst: 1.050.000 tokena. Maks. izlaz: 128K tokena. API cena: $30 ulaz / $180 izlaz po 1M tokena standardno. Za upite iznad 272K tokena: 2x ulaz i 1,5x izlaz. Regionalni serveri: 10% uvećanje. Benčmarkovi: GPQA Diamond 94,4; ARC-AGI-2 83,3; Humanity's Last Exam 58,7 sa alatima; BrowseComp 89,3. Neki API zahtevi mogu trajati minute — preporučuje se pozadinski mod.
Najpogodnije za: Izuzetno teški problemi rezonovanja, visokoetažni pregled, slučajevi gde cena nije važna koliko kvalitet odgovora
GPT-5.4 mini — manji model za rezonovanje
Objavljen: 17. mart (API/Codex), 18. mart (ChatGPT) | API: $0,75/$4,50 po 1M tokena
Jeftiniji API model i praktičan rezervni put u ChatGPT korisničkom iskustvu. Za Free i Go korisnike pojavljuje se kroz Thinking funkcionalnost. Za većinu plaćenih korisnika to je model na koji ChatGPT prelazi kada se dostigne limit GPT-5.4 Thinking-a.
★ Šta je novoLansiran 17–18. marta 2026. kao opcija Thinking za Free/Go korisnike i kao rezervni model za plaćene korisnike koji dostignu limite GPT-5.4 Thinking-a.
Tehnički detaljiAPI ID: gpt-5.4-mini. Kontekst: 400K tokena. API cena: $0,75 ulaz / $0,075 keš / $4,50 izlaz po 1M tokena. Batch: pola standardne. Benčmarkovi: SWE-bench Pro 54,4; Terminal-Bench 60,0; MCP Atlas 57,7.
Najpogodnije za: Rezonovanje koje vodi računa o budžetu, graciozni prelaz po dostizanju limita, API radni procesi velikog obima koji i dalje zahtevaju snažan model
GPT-5.4 nano — najjeftiniji model samo za API
Objavljen: 17. mart 2026. | API: $0,20/$1,25 po 1M tokena | Samo API
Nije mainstream ChatGPT model. Postoji za developere kojima je potreban izuzetno jeftin i brz model za automatizaciju, klasifikaciju, ekstrakciju i pozadinske zadatke velikog obima.
★ Šta je novoLansiran 17. marta 2026. kao najjeftiniji trenutni API model GPT-5.4 klase.
Tehnički detaljiAPI ID: gpt-5.4-nano. Kontekst: 400K tokena. API cena: $0,20 ulaz / $0,02 keš / $1,25 izlaz po 1M tokena. Benčmarkovi: SWE-bench Pro 52,4; Terminal-Bench 46,3; MCP Atlas 56,1.
Najpogodnije za: Developer automatizacija, klasifikacija, ekstrakcija, procesi zaključivanja koji vode računa o troškovima
Nasleđeni modeli — šta ostaje i do kada
GPT-5.2 Thinking ostaje u Nasleđenim modelima za Plus i Pro korisnike 90 dana od lansiranja 5. marta 2026. — povlači se 5. juna 2026. GPT-4o u Custom GPT-ovima: Business, Enterprise i Edu ga zadržavaju do 3. aprila 2026. Već povučeno iz ChatGPT-a (13. feb. 2026.): GPT-4o, GPT-4.1, GPT-4.1 mini, OpenAI o4-mini i GPT-5 (Instant i Thinking).
Agent Mode — ChatGPT koji može da dela
Široko dostupno od marta 2026. | Uključeno u Plus, Pro, Business, Enterprise, Edu
Umesto da samo odgovara savetima, Agent Mode može pregledati veb-sajtove, raditi kroz formulare, pregledati fajlove, koristiti alate i obavljati višekoračne zadatke. Kada sajt zahteva prijavu ili osetljivu potvrdu, kontrola se vraća korisniku. Tipično trajanje zadatka: 5–30 minuta.
Tehnički detaljiLimiti: Plus 40 agent poruka/mesec; Pro 400/mesec; Business/Enterprise 40/mesec. Business fleksibilno određivanje cena: 30 kredita po poruci. Korisnik ostaje u kontroli za prijave — snimci ekrana se ne prave dok korisnik kontroliše pregledač.
Najpogodnije za: Delegiranje dugih veb zadataka, repetitivno prikupljanje informacija, rad sa mnogo formulara
Canvas — deljeni prostor za uređivanje
Dostupnost: veb, Windows, macOS; iOS/Android uskoro | Nije dostupno sa GPT-5.4 Pro
Trajni canvas gde se tekst ili kod može revidirati, anotirati i iterirati direktno. Podržava izvršavanje Python koda i može biti omogućen unutar GPT-ova. Najkorisnije kada zadatak postane previše velik za obični razgovor.
Najpogodnije za: Dugi nacrti, iterativno uređivanje, strukturirani pregled koda
Sora 2 — generisanje video zapisa
Lansiran: 30. septembra 2025. | Dostupno na sora.com, iOS, Android i ChatGPT planovima
OpenAI-jev trenutni sistem za generisanje video zapisa. Limiti za potrošače: Plus/Business — neograničene slike i video, do 480p i 10 sekundi, 1 istovremeno generisanje. Pro — neograničeno, brže, do 1080p i 20 sekundi, do 5 istovremenih, preuzimanja bez vodenog žiga.
★ Šta je novo12. mart 2026.: Sora API proširen sa ponovo upotrebljivim referencama likova, duljim generisanjem od 20 sekundi, 1080p izlazom za sora-2-pro, proširenjima video zapisa i podrškom za Batch API.
Tehnički detaljiAPI ID-ovi: sora-2 i sora-2-pro. API cena: sora-2 na 720p je $0,10/sek. standardno. sora-2-pro: $0,30/sek. na 720p, $0,50/sek. na 1024p, $0,70/sek. na 1080p; batch je pola standardnih cena.
Najpogodnije za: Kreativno generisanje video zapisa, vizualizacija koncepta, marketinški nacrti, storyboarding
Deep Research — dugoformno istraživanje s citatima
Dostupno u ChatGPT-u | Izlaz: Markdown, Word ili PDF
Kreira plan istraživanja, prolazi kroz izvore tokom vremena i vraća strukturirani izveštaj s citatima. Korisnik može uređivati plan istraživanja pre nego što počne izvršavanje. Nasleđeni Deep Research mod uklonjen je 26. marta 2026.
★ Šta je novo19. mart 2026.: OpenAI dodao uređivanje plana istraživanja i poboljšan prikaz izveštaja. Nasleđeni Deep Research mod uklonjen 26. marta 2026.
Najpogodnije za: Politički brifovi, istraživanje tržišta, pregledi poput literature, sinteza iz više izvora
Memorija, Projekti i Biblioteka — sloj persistencije
Memorija i Projekti aktivni | Biblioteka lansirana 23. marta 2026.
Memorija omogućava ChatGPT-u da pamti prošle detalje. Projekti grupišu razgovore i fajlove zajedno. Biblioteka je novo mesto za čuvanje uploadovanih i generisanih fajlova — ponovo upotrebljivo između razgovora.
★ Šta je novo23. mart 2026.: lansirana Biblioteka za sačuvane fajlove. Trenutno samo veb; nedavni fajlovi u kompozeru i pretraga fajlova na iOS-u i Android-u. Dostupno za Plus, Pro i Business korisnike van EEA, Švajcarske i UK.
Najpogodnije za: Tekući projekti, ponavljajući rad, personalizovana asistencija
OpenAI Codex — platforma za kodiranje
Aplikacija lansirana: 2. februara 2026. | Uključeno u Plus, Pro, Business, Edu, Enterprise
Namenski prostor za agente kodiranja sa više paralelnih agenata, izolovanim radnim stablima, pregledivim razlikama i delegiranjem cloud agentima. Integracije: GitHub, Slack, Linear.
★ Šta je novo4. mart 2026.: prošireno na Windows za Business prostore. 10. mart: automatsko dopunjavanje za deljene kredite koje koriste Codex i Sora. GPT-5.3-Codex ostaje specijalizovani model za kodiranje lansiran 5. februara 2026.
Najpogodnije za: Ozbiljan softverski rad, paralelni zadaci kodiranja, timovi koji žele agentsku pomoć u pravim razvojnim procesima
Planovi i cene
| Plan |
Cena |
Šta uključuje |
| Free |
$0 |
Ograničen GPT-5.3, ograničeni upload i generisanje slika, ograničen Deep Research |
| Go |
$8/mes. (SAD) |
Više GPT-5.3, više upload-a i slika, dulja memorija. Može uključivati reklame. |
| Plus |
$20/mes. |
Napredni modeli rezonovanja, projekti, zadaci, Codex, Sora, prošireni Deep Research i Agent Mode |
| Pro |
$200/mes. |
GPT-5.4 Pro, maksimalni Deep Research i Agent Mode, proširena Sora, Codex višeg prioriteta |
| Business (ranije Team) |
$25/korisnik/mes. godišnje |
Deljeni radni prostor, admin kontrole, Codex, pristup agentu. Min 2 korisnika. |
| Enterprise |
Po dogovoru |
Veći kontekst/fajl podrška, SCIM, RBAC, EKM, rezidencija podataka, SLA |
Tehnički detaljiAPI cena (trenutna GPT-5 porodica): gpt-5.4 $2,50/$0,25 keš/$15 izlaz po 1M tokena; gpt-5.4-pro $30/$180; gpt-5.4-mini $0,75/$0,075/$4,50; gpt-5.4-nano $0,20/$0,02/$1,25. API naplata je odvojena od ChatGPT pretplata. Samouslužni API nivoi korišćenja skaliraju od ~$100/mes. na nižim nivoima do $200.000/mes. na nivou 5.
Gemini (Google)
Gemini je Google-ov primarni sistem veštačke inteligencije — jedinstveni "mozak" koji može videti, čuti, čitati i kreirati. Multimodalan je: razume tekst, slike, audio i video odjednom. Pokreće sve, od jednostavnog čet boksa na vašem telefonu do profesionalnih alata koje koriste naučnici i programeri.
Gemini 3.1 Flash Lite — specijalista za efikasnost
3. mart 2026. | $0,25 / 1M tokena | Programeri i preduzeća
Najnoviji i najpristupačniji član porodice. Napravljen za rad velikog obima gde treba da AI obavlja jednostavan zadatak hiljadama puta u sekundi bez velikih troškova.
★ Šta je novoLansiran 3. marta 2026. kao prvi "Lite" model u 3.1 generaciji, optimizovan za produkcijske procese osetljive na kašnjenje. 2,5x brže vreme do prvog tokena (TTFT) od Gemini 2.5 Flash-a.
Tehnički detaljiKontekstni prozor: 1M tokena (32K putem API Preview). API ID: gemini-3.1-flash-lite-preview. API cena: $0,25 po 1M ulaz / $1,50 po 1M izlaz. Benčmarkovi: GPQA Diamond: 86,9% | MMMU-Pro: 76,8% | MMMLU: 88,9%.
Najpogodnije za: Obrada podataka velikog obima, produkcijski procesi osetljivi na kašnjenje
Gemini 3.1 Flash — uravnoteženi radni konj
Februar 2026. | $0,50 / 1M tokena | Svi korisnici (Free i plaćeni)
Podrazumevana verzija Gemini-ja za većinu ljudi. Postiže najboljiu ravnotežu između inteligentnosti i brzine. Upravljačka planiranjem putovanja, sažimanjem PDF-ova, pisanjem e-pošte i razgovorom u realnom vremenu.
Tehnički detaljiKontekstni prozor: 1M tokena. API cena: $0,50 po 1M ulaz / $3,00 po 1M izlaz. API ID: gemini-3.1-flash. Benčmarkovi: ARC-AGI-2: 45,2% | MMLU-Pro: 82,1%.
Najpogodnije za: Opšti čet i brzi svakodnevni zadaci
Gemini 3.1 Pro — flagship model
19. februar 2026. | $2,00 / 1M tokena | AI Pro i Ultra pretplatnici
Google-ov najsposobniji model za rezonovanje. Dizajniran za teške probleme: debagovanje složenog koda, sinteza konfliktnih podataka iz više izvora, kreiranje zamršenih živih kontrolnih tabli ili animiranih grafika. Sadrži "Prošireno razmišljanje" — pauzira da razmisli pre nego što odgovori na teške logičke zagonetke.
Tehnički detaljiKontekstni prozor: 1M do 2M tokena. Izlazni prozor: 65.536 tokena. API cena: $2,00 po 1M ulaz / $12,00 po 1M izlaz (do 200k konteksta); cene se udvostručuju iznad 200k. Benčmarkovi: ARC-AGI-2: 77,1% (SOTA) | GPQA Diamond: 94,3% | Terminal-Bench 2.0: 68,5%.
Najpogodnije za: Složena logika i kodiranje, naučno istraživanje, sinteza iz više izvora
Gemini 3.1 Ultra — najmoćniji model
Februar 2026. | Samo pretplata ($249,99/mes.) | AI Ultra pretplatnici
Vrhunski model, ekskluzivan za najskuplji plan. Koristi se za najzahtevnije kreativne zadatke: generisanje profesionalnog 4K video zapisa sa Veo 3.1, pokretanje naprednih Deep Research izveštaja koji mogu trajati do 20 minuta.
Tehnički detaljiTrenutno drži #1 na Artificial Analysis Intelligence Index (rezultat: 61). API pristup ograničen — primarno putem Vertex AI za preduzeća i kroz Gemini App za Ultra pretplatnike.
Najpogodnije za: Vrhunsko istraživanje, generisanje 4K video zapisa, najzahtevniji kreativni rad
NotebookLM — personalizovani istraživački asistent
Učitajte sopstvene dokumente (PDF-ove, transkripte, veb-sajtove) i NotebookLM kreira privatni svet gde AI odgovara isključivo na osnovu vaših specifičnih informacija. Audio pregledi: dva AI "domaćina" razgovaraju o vašim dokumentima kao u podkastu. Studijska oprema: kvizovi, fleš kartice i studijski vodiči iz vaših izvora.
★ Šta je novoKinematografski video pregledi (15. mart 2026.): lansiran za Ultra korisnike za kreiranje 4K vizuelnih sažetaka. Mape uma i Artifakti: automatski generiše vizuelne dijagrame koji prikazuju kako su vaše ideje povezane.
Tehnički detaljiLimiti za izvore: do 50 izvora po beležnici; 500k reči po izvoru. Izvoz: podržava kreiranje .pptx i .docx artifakata.
Najpogodnije za: Istraživanje i učenje, Q&A nad privatnim bazama znanja
Jules — AI agent za kodiranje
Asinhronog partnera za kodiranje. Date mu veliki zadatak, zatvorite laptop, a Jules radi u pozadini — klonira vaš kod, testira sopstveni rad i šalje vam zahtev za povlačenje (pull request) kada završi.
Tehnički detaljiBazni model: Gemini 2.5 Pro / 3.1 Pro. Limiti: Free (15 zadataka/dan), Pro (100 zadataka/dan), Ultra (300 zadataka/dan).
Najpogodnije za: Autonomno kodiranje, izvršavanje zadataka u pozadini
Veo 3.1 i Lyria 3
- Veo 3.1: Google-ov model za generisanje 4K video zapisa. Kreira 8-sekundne kinematografske isečke sa sinhronizovanim zvukom.
- Lyria 3: Visokoverna muzička generacija. Kreira 3-minutne pesme sa vokalom i tekstovima na osnovu tekstualnog upita.
★ Šta je novoLyria 3 (25. mart 2026.): ažuriran da podržava 3-minutne profesionalne aranžmane sa realističnim vokalima.
Planovi i cene
| Plan |
Cena |
Šta uključuje |
Nije uključeno |
| Basic (Free) |
$0,00 |
Gemini 3.1 Flash, 32k kontekst |
Deep Research, Jules, 4K video |
| AI Plus |
$7,99/mes. |
200GB skladišta, 200 kredita |
3.1 Pro model, visoki limiti za Jules |
| AI Pro |
$19,99/mes. |
2TB skladišta, Gemini 3.1 Pro |
Veo 3.1 Standard (4K) |
| AI Ultra |
$249,99/mes. |
30TB skladišta, Gemini 3.1 Ultra |
— |
Microsoft Copilot
Microsoft Copilot nije jedinstven AI model. To je porodica AI funkcija zasnovanih na modelima iz drugih kompanija (primarno OpenAI GPT-5 i Anthropic Claude). Ono što Copilot čini posebnim je Work IQ, integracija sa Microsoft Graph-om i mogućnost da boravi unutar Office aplikacija kao što su Word, Excel i Teams. Osnovni modeli obezbeđuju sposobnosti sirovog rezonovanja i jezičke obrade, dok Copilot dodaje organizacioni kontekst, usklađenost i integraciju produktivnosti.
Copilot Free
Cena: besplatno
Osnovna asistencija za čet i pisanje unutar Microsoft Edge-a i Office veb aplikacija. Ograničeno na upite opšte namene i ne pristupa organizacionim podacima.
Tehnički detaljiKoristi GPT-4.5. Bez Microsoft Graph integracije. Bez okvira za usklađenost. Bez admin kontrola.
Najpogodnije za: Individualni korisnici koji žele laganu AI pomoć bez enterprise integracije
Copilot Pro
Cena: $20/mes.
Otključava napredne funkcionalnosti pisanja, sažimanja i dizajna u Word-u, Excel-u, PowerPoint-u i Outlook-u. Zahteva posebnu Microsoft 365 pretplatu za punu integraciju aplikacija.
Tehnički detaljiPristup GPT-5 i Claude 3. Ograničena Graph integracija. Bez enterprise usklađenosti. Admin kontrole nisu dostupne.
Najpogodnije za: Power korisnici i slobodnjaci koji trebaju dublju AI integraciju u Office aplikacijama
Copilot Business
Cena: $30/korisnik/mes.
Integracija sa Microsoft Graph-om omogućuje kontekstnu asistenciju kroz Teams, SharePoint i OneDrive. Uključuje transkripciju sastanaka, AI beleške i prevod u živo.
Tehnički detaljiGraph integracija: SharePoint, OneDrive, Teams. Usklađenost: ISO 27001, SOC 2. Admin portal sa kontrolama zasnovanim na ulogama.
Najpogodnije za: Mala i srednja preduzeća kojima su potrebne funkcionalnosti saradnje i usklađenosti
Copilot Enterprise
Cena: $30/korisnik/mes. (zahteva E3/E5 licencu)
Dodaje Semantic Index, Purview upravljanje, API-je za usklađenost, SCIM provizioniranje i vertikalne dodatke kao što su Copilot for Sales i Service.
Tehnički detaljiPuna Graph integracija. Semantic Index za organizaciono znanje. Purview usklađenost. SCIM podrška. API pristup za enterprise programere.
Najpogodnije za: Velike organizacije koje zahtevaju usklađenost, upravljanje i proširivost
E7 nivo
Cena: $57/korisnik/mes.
Kombinuje Microsoft 365 E7 sa Copilot Enterprise. Uključuje naprednu bezbednost, usklađenost i analitiku.
Najpogodnije za: Preduzeća koja standardizuju na Microsoft-ovom najvišem nivou bezbednosti i usklađenosti
Agent 365
Cena: $15/korisnik/mes.
Lagan okvir agenta za automatizaciju radnih tokova. Omogućuje Copilot-u da deluje kroz aplikacije i servise bez pune Microsoft 365 integracije. Agent zasnovan na Claude 3.
Najpogodnije za: Timovi kojima je potrebna automatizacija bez punog enterprise licenciranja
Copilot Cowork
Cena: Istraživački pregled
Kolaborativni AI agent izrađen sa Anthropic-om. Naglašava Work IQ — sposobnost praćenja zadataka, kontrolnih tačaka i petlji od plana do akcije.
★ Šta je novo18. mart 2026.: Cowork dodao svesnost o pristiglim porukama i kalendaru u Copilot Chat-u.
Tehnički detaljiClaude Cowork arhitektura. Graph integracija opcionalna. Usklađenost: SOC 2, ISO 27001. Enterprise Cowork dodaje API-je za upravljanje.
Najpogodnije za: Timovi koji eksperimentišu sa kolaborativnim AI procesima
GitHub Copilot
Cena: Besplatno (studenti/OSS), Pro $10/mes., Pro+ $15/mes., Business $19/korisnik/mes., Enterprise $39/korisnik/mes.
Pomaže pri dovršavanju koda, rezonovanju kroz više fajlova i dokumentaciji. Enterprise planovi dodaju prilagođene baze znanja i fino podešene modele.
Tehnički detaljiAgent za više fajlova. Enterprise podržava fino podešavanje privatnih modela. Usklađenost: SOC 2, ISO 27001.
Najpogodnije za: Programeri i preduzeća kojima je potrebna AI asistencija za kodiranje
Copilot unutar Microsoft 365 aplikacija
- Word: Pisanje nacrta, sažimanje, prepisivanje
- Excel: Analiza podataka, generisanje formula, kreiranje grafikona
- PowerPoint: Generisanje slajdova, predlozi dizajna
- Outlook: Pisanje e-pošte, sažimanje
- Teams: Transkripcija sastanaka, AI beleške, prevod u živo
- SharePoint: Sažimanje sadržaja, označavanje metapodataka
- OneDrive: Sažimanje fajlova, asistencija pri pretrazi
★ Šta je novo (mart 2026.)Excel: dodata prediktivna analiza trendova.
Najpogodnije za: Korisnici Microsoft 365-a koji žele AI direktno ugrađen u aplikacije za produktivnost
Planovi i cene
| Nivo |
Cena |
Zahtevi |
Modeli |
| Copilot Free |
$0 |
Microsoft nalog |
GPT-4.5 |
| Copilot Pro |
$20/mes. |
Microsoft 365 pretplata |
GPT-5, Claude 3 |
| Copilot Business |
$30/korisnik/mes. |
Min. 5 sedišta |
GPT-5, Claude 3 |
| Copilot Enterprise |
$30/korisnik/mes. |
E3/E5 licenca |
GPT-5, Claude 3 |
| E7 nivo |
$57/korisnik/mes. |
Enterprise licenca |
GPT-5 |
| Agent 365 |
$15/korisnik/mes. |
Nema posebnih zahteva |
Claude 3 |
| GitHub Copilot Pro |
$10/mes. |
GitHub nalog |
GPT-4.5 |
| GitHub Copilot Enterprise |
$39/korisnik/mes. |
Enterprise licenca |
GPT-5 |
Ovaj dokument je ažuriran na dan marta 2026. Za najnovije podatke, zatražite novo pokretanje ovog vodiča mesečno.
