agents
Best AI Agents
Compare the best AI agents in 2026 — ChatGPT Agent, Claude Code, Google Antigravity, Grok Build, Manus and Devin — plus the agentic models (Opus 4.8, GPT-5.5, Fable 5) and benchmarks (OSWorld, τ²-bench, SWE-bench Pro) that power them, with decisive picks for every use case.
Quick answer: “Best AI agent” splits three ways in mid-2026. For a general agent that browses, fills forms and finishes online tasks for you, ChatGPT Agent (in ChatGPT Plus, $20/month) is the most capable all-rounder. For agentic coding — the most mature category — Claude Code paired with Claude Opus 4.8 is the developer favourite, with Grok Build and Google Antigravity 2.0 as the fastest-moving challengers. For a do-it-all autonomous “super agent” that produces finished research, slides and spreadsheets, Manus leads. The strongest underlying agentic model you can run is Anthropic’s restored Claude Fable 5, with GPT-5.5 the leader on tool-use and terminal work. The caveat that frames everything below: agent benchmarks are unreliable — in April 2026 researchers showed all eight major ones could be gamed to near-perfect scores without solving a single task (UC Berkeley RDI) — and Gartner expects over 40% of agentic-AI projects to be cancelled by 2027 (Gartner).
The honest answer depends on what you want the agent to do: run your computer, write your code, automate a business process, or be built into your own product. This guide covers the whole stack — the models that give agents their reasoning, the consumer and coding agents you can use today, the enterprise platforms, and the developer frameworks for building your own — with current benchmarks, pricing and real sentiment. Two things define the 2026 moment: the tooling has genuinely crossed from demo to daily use, and the gap between the marketing and the measured reliability has never been wider.
The current state of AI agents: July 2026
An AI agent is a model given tools, memory and a goal, allowed to take multiple steps on its own — browsing, running code, calling APIs, editing files — until a task is done. In 2026 that shifted from a research demo to something people use daily, but adoption and trust are moving in opposite directions.
Gartner projects that 40% of enterprise applications will embed task-specific agents by the end of 2026, up from under 5% in 2025 (Gartner). Yet only about 17% of organisations have actually deployed an agent, with 42% expecting to within a year (Gartner via Joget). The global agent market is estimated at $10.9–12.1 billion in 2026, growing 44–46% a year. The demand is real; the delivered value is patchy.
Five shifts define where the category sits now.
-
The benchmarks broke. On 12 April 2026, UC Berkeley’s Center for Responsible Decentralized Intelligence published research showing an automated scanning agent could reward-hack all eight major agent benchmarks — SWE-bench, WebArena, OSWorld, τ²-bench and others — reaching near-perfect scores without completing the underlying tasks (benchmarkingagents.com). Every headline agent number should now be read with that in mind: vendor claims are a ceiling, standardised harnesses a floor, and both are gameable.
-
“Agent washing” got called out. Gartner estimates only about 130 of the thousands of vendors marketing “agentic AI” ship anything genuinely agentic; the rest are rebranded chatbots, RPA and assistants (Gartner). The same analysis forecasts 40%+ of agentic projects cancelled by 2027 on cost, unclear value and weak risk controls.
-
Agentic coding matured into a three-way race. The most reliable agents today write code. The category consolidated around Claude Code (the developer favourite), OpenAI’s Codex CLI, and xAI’s new Grok Build — with Google’s Antigravity 2.0 platform and Cognition’s Devin pushing full autonomy (CIO Dive).
-
The frontier model behind agents came back. Anthropic’s Mythos-class Fable 5 — the strongest agentic model on raw benchmarks — was suspended worldwide from 12 June under a US export-control directive, then returned to general availability on 1 July 2026 once the controls were lifted (Anthropic). The practical default remains Opus 4.8; the ceiling is usable again.
-
Building agents got a real toolchain. Every major lab now ships an agent SDK — Anthropic’s Claude Agent SDK, OpenAI AgentKit and Agents SDK, Google’s Agent Development Kit and Managed Agents — alongside framework leaders LangGraph and CrewAI. Search demand for “claude agent sdk” alone rose from 50 to 14,800 monthly queries in a year (Morph). The unit of software is shifting from an app you use to an agent you delegate to.
Top AI agent models (July 2026)
An agent is only as good as the model reasoning underneath it. Three benchmark families matter for agents, and they measure different things: SWE-bench Pro (resolving real software issues), OSWorld (operating a real computer — clicking, typing, navigating apps), and τ²-bench (using tools to complete customer-service tasks while following a written policy). No single model wins all three.
| # | Model | Provider | Tier | OSWorld (computer use) | Terminal-Bench 2.1 | SWE-bench Pro | Price (in/out per MTok) |
|---|---|---|---|---|---|---|---|
| 1 | Claude Fable 5 | Anthropic | Frontier | 79.6%† | n/a | 80.3%ᵖᵛ | $10 / $50 |
| 2 | Claude Mythos 5 | Anthropic | Frontier | 79.6%† | n/a | 80.3%ᵖᵛ | $10 / $50 |
| 3 | Claude Opus 4.8 | Anthropic | Flagship | 72.7%‡ | 74.6% | 69.2%ᵖᵛ | $5 / $25 |
| 4 | GPT-5.6 Sol | OpenAI | Flagship | n/a | leads OpenAI harness | n/a | $5 / $30 |
| 5 | GPT-5.5 | OpenAI | Flagship | 78.7%ᵛ | 78.2% | 58.6%ᵖᵛ | $5 / $30 |
| 6 | Gemini 3.5 Pro | Flagship | n/a | n/a | n/a | Limited | |
| 7 | Grok 4.3 | xAI | Flagship | n/a | n/a | n/a | $1.25 / $2.50 |
| 8 | GPT-5.4 | OpenAI | Strong | 75.0% | n/a | 59.1%ᵖˢ | $2.50 / $15 |
| 9 | Gemini 3.1 Pro | Strong | n/a | n/a | 46.1%ᵖˢ | $2 / $12 | |
| 10 | Gemini 3.5 Flash | Value | n/a | 76.2% | n/a | n/a | |
| 11 | MiniMax M3 | MiniMax | Open | n/a | 66.0% | 59.0%ᵖ | $0.30 / $1.20 |
| 12 | GLM-5.2 | Zhipu | Open | n/a | n/a | 62.1%ᵖᵛ | Open / self-host |
Reading the columns: ᵖᵛ = SWE-bench Pro on the vendor’s harness (a ceiling); ᵖˢ = SWE-bench Pro on Scale’s standardized harness (a floor); ᵛ = the Verified/OSWorld-Verified variant. † = Mythos Preview’s May 2026 OSWorld result, the closest published proxy for the Mythos-class family. ‡ = Opus 4.6’s OSWorld result; Anthropic has not separately published a 4.8 figure. n/a means we don’t have a verified figure for that cell, usually because the model page or an independent run is still pending — not that the capability is absent.
What the numbers say — and don’t
GPT-5.5 is the tool-use and terminal leader. It tops Terminal-Bench 2.1 (78.2%), leads OSWorld-Verified (78.7% to Claude’s 78.0%), and hits 98.0% on τ²-bench Telecom with no prompt tuning — the most enterprise-relevant agent score published (decodethefuture). If your agent’s job is calling tools and running commands, GPT-5.5 (and the GPT-5.6 Sol preview above it) is the model to beat.
Anthropic leads long-horizon and coding agents. Fable 5 posts the highest coding-agent scores of any model (80.3% SWE-bench Pro, vendor), and Anthropic swept the top six places on the GAIA general-assistant leaderboard through April 2026 (awesomeagents.ai). For everyday building, Opus 4.8 is the pragmatic pick — half Fable 5’s price, no tightened safety classifier, and markedly more honest about flagging its own mistakes.
Computer-use agents have reached human range — barely. The OSWorld human baseline is 72–84% depending on task (decodethefuture). Frontier models now sit inside that band on average, but still fail the ~20% of tasks needing fine motor control or unfamiliar menus. Treat “human-level computer use” as true for routine flows and false for edge cases.
The open-weight tier is closing on agents too. MiniMax M3 (Terminal-Bench 66.0%, BrowseComp 83.5) and GLM-5.2 (built for long-horizon agents, MIT-licensed) bring genuinely useful agentic capability to self-hosted stacks at a fraction of the cost — see the open tier of our best AI models ranking.
Why you should distrust every agent benchmark
The scores above are useful signposts, not guarantees. Beyond the harness variance that already splits SWE-bench Pro by 17–21 points between vendor and standardized runs, the UC Berkeley reward-hacking finding (April 2026) showed all eight leading agent benchmarks can be gamed to near-perfect scores without doing the task (benchmarkingagents.com). The practical rule: weight a model’s behaviour on your workflow over any leaderboard, run a small private eval before you commit, and treat a two-point benchmark gap as noise.
Best AI agent tools and platforms compared
We’ve grouped the field by what you actually want the agent to do, and ordered within each group by current standing.
1. ChatGPT Agent — the best general-purpose agent
Price: Included with ChatGPT Plus ($20/month) and Pro ($200/month); not on the free tier Runs on: GPT-5.5, inside the ChatGPT app and web Best for: Everyday online tasks — research, bookings, forms, spreadsheets
ChatGPT Agent is the most-used general agent because it’s built into the most-used app. Invoke it with /agent and it reasons, browses websites, works with your uploaded files, connects to email and document sources, fills out forms and edits spreadsheets — pausing for confirmation on anything consequential (OpenAI). It’s the safe default for a non-technical user who wants a computer-using assistant without wiring anything up.
Limitations: Gated to paid tiers; can be slow on long tasks; and like all browser agents it still stumbles on unusual site layouts. Best for delegated web work, not for autonomous coding.
2. Claude Code (+ Claude Agent SDK) — best for agentic coding and builders
Price: Included with Claude Pro ($20/month) and Max ($100–200/month); API/credit billing for programmatic use Runs on: Opus 4.8, Sonnet 4.6, Haiku 4.5 Best for: Agentic software work; building custom production agents
Claude Code is the developer-favourite agentic coding tool (46% “most-loved” in JetBrains’ 2026 survey) and pairs the best available coding model with a strong native harness — see our best AI for coding guide for the full picture. Underneath it, the Claude Agent SDK exposes the same primitives — file editing, bash execution, web search/fetch, subagents, persistent sessions, MCP support and human-in-the-loop checkpoints — so you can build production agents on the harness that powers Claude Code (Anthropic). In 2026 it became the default way developers wire Claude into long-running workflows; a separate credit pool for programmatic use went live on 15 June 2026.
Limitations: Model choice is Claude-only. Heavy Opus use adds up — route routine steps to Haiku 4.5.
3. Google Antigravity 2.0 — best multi-agent platform
Price: Free tier; Pro $20/month; Ultra $100/month Runs on: Gemini 3.x, with multi-agent orchestration Best for: Running and observing several agents in parallel; Gemini-ecosystem teams
Google Antigravity 2.0 launched at Google I/O on 19 May 2026 as a standalone, agent-first platform — desktop app, CLI, SDK and a managed agent service in one release (TechCrunch). Its Manager Surface lets you spawn and watch multiple agents at once, and a Browser Subagent tests in a real browser. It also absorbed the retired Project Mariner (shut down 4 May 2026), folding Google’s computer-use lineage into the platform, and Managed Agents in the Gemini API now spin up a tool-using agent in an isolated Linux environment with one call (Google).
Limitations: Gemini-only models; early reviews flag quota limits on cheaper tiers and the usual churn of a fast-moving platform.
4. Grok Build — xAI’s agentic coding CLI
Price: Beta included with SuperGrok ($30/month) and X Premium Plus; launched to SuperGrok Heavy ($300/month) first Runs on: grok-build-0.1 (256K-token context) Best for: Terminal-native coding for xAI-ecosystem developers
Grok Build is xAI’s entry into the coding-agent race — a terminal CLI announced in 2026 that turns natural language into code and automation. It runs a plan mode (approve, comment on or rewrite the plan before execution) and delegates larger jobs to up to eight parallel subagents, each working a plan-search-build loop (Engadget). Its 256K context holds a mid-sized codebase in memory. It makes the coding-agent field a genuine three-way contest with Claude Code and Codex.
Limitations: New and beta-stage; smaller community and ecosystem than Claude Code or Codex; tied to the Grok/X subscription stack.
5. Manus — best autonomous “super agent”
Price: Free (300 daily credits, Manus 1.6 Lite); Standard $20/month (4,000 credits); Customizable $40/month (8,000); Extended $200/month (40,000); Team from $20/seat Runs on: Orchestrates frontier models from other labs Best for: End-to-end deliverables — research reports, slide decks, data tasks
Manus is the leading do-it-all autonomous agent app: describe an outcome and it plans, browses, runs code and returns a finished deliverable. It orchestrates other labs’ frontier models rather than shipping its own, and prices by credits — every browse, code run or file analysis consumes them, and monthly credits don’t roll over (No Code MBA). Meta’s attempted $2bn acquisition was blocked by China’s antitrust regulator in April 2026, leaving ownership unresolved.
Limitations: Credit costs are hard to predict on complex jobs; because it orchestrates external models, output quality tracks whatever it routes to.
6. Devin — most autonomous software engineer
Price: Core pay-as-you-go from $20 (compute units ~$2.25 each); Team $500/month (250 units at ~$2.00); Enterprise custom Runs on: Cognition’s own stack in a sandboxed cloud environment Best for: High-volume, well-scoped engineering tasks delegated like tickets
Devin, from Cognition, is the most fully autonomous coding agent: it reads tickets from Linear, Jira or Slack, plans, writes the implementation, runs tests and opens a PR — the whole loop unattended. A February 2026 update added parallel sessions and better long-task context retention. Nubank reported 8–12x efficiency gains using it on a 6-million-line migration (Idlen). Note the separate Devin Desktop — the IDE relaunched from Windsurf — covered in our coding guide.
Limitations: Strong on bounded, repetitive work; weak and expensive on novel architecture or ambiguous requirements — independent 2026 reviews land it around 7.5/10 with frequent human intervention needed.
7. Genspark — best for finished multi-step deliverables
Price: Free tier; paid plans (credit-based) Runs on: Orchestrates 30+ models Best for: “Super agent” outputs — slides, research, even automated phone calls
Genspark is a reported $1.25bn “super agent” that orchestrates 30+ models to produce done-for-you deliverables — formatted slide decks, research reports, and outbound phone calls. It sits alongside assistants like ChatGPT rather than replacing them, and competes most directly with Manus for the autonomous-deliverables use case.
Limitations: Like Manus, output depends on routing and credits can climb; narrower general-chat ability than the first-party assistants.
8. Perplexity Computer — best multi-model orchestration
Price: On Perplexity’s paid tiers (Pro $20/month; Max $200/month) Runs on: Coordinates 19+ specialised models Best for: Long-running research and workflow tasks that span many models
Launched in February 2026, Perplexity Computer coordinates 19+ specialised models to execute long-running workflows, routing each subtask to the best-suited model (Salesforce roundup). It extends Perplexity’s sourced-answer strength from single questions to multi-step jobs, and is a natural fit for anyone who already lives in Perplexity for research.
Limitations: Best on research-shaped work; not a general desktop-automation or coding agent.
9. Salesforce Agentforce — enterprise leader
Price: Usage-based enterprise pricing (contact sales) Runs on: Atlas Reasoning Engine over your Salesforce data Best for: Customer-facing and internal agents grounded in CRM data
Salesforce Agentforce is the most-deployed enterprise agent platform, with a reported 18,500+ deals across 12,500+ companies in 39 countries. Its Atlas Reasoning Engine grounds agents in your real-time CRM metadata and business logic, with a trust layer for data masking and retention controls. For companies already on Salesforce, it’s the shortest path to production agents that act on customer data.
Limitations: Most valuable inside the Salesforce ecosystem; pricing and governance need careful modelling for high-volume use.
10. Microsoft Copilot Studio — best low-code enterprise builder
Price: Consumption-based within Microsoft 365 Runs on: OpenAI and Microsoft models over Microsoft Graph Best for: Building agents inside Teams, SharePoint and Dynamics without code
Microsoft Copilot Studio is the low-code path for the ~90% of large enterprises already on Microsoft 365. It builds, customises and deploys agents grounded in your organisation’s own data through Microsoft Graph — email, documents, Teams — and embeds them across the Microsoft stack. It’s the enterprise default where governance and existing tenancy matter more than raw model choice.
Limitations: Real value is concentrated inside the Microsoft ecosystem; consumption pricing needs forecasting for heavy workloads.
Feature comparison: the full matrix
| Agent | Category | Autonomy | Model(s) | Interface | Headline price | Best for |
|---|---|---|---|---|---|---|
| ChatGPT Agent | General | High (checkpoints) | GPT-5.5 | ChatGPT app/web | $20/mo | Everyday online tasks |
| Claude Code | Coding | Full | Claude only | CLI, desktop, IDE | $20–200/mo | Agentic coding |
| Antigravity 2.0 | Platform | Full (multi-agent) | Gemini 3.x | Desktop/CLI/SDK | $0 / $20 / $100 | Parallel agents |
| Grok Build | Coding | Full (8 subagents) | grok-build-0.1 | Terminal CLI | $30/mo (beta) | Terminal coding |
| Manus | Super agent | Full | Multi (orchestrated) | Web app | $0–200/mo | Finished deliverables |
| Devin | Coding | Full (unattended) | Cognition stack | Cloud + Slack | From $20 (usage) | Ticket-scoped eng |
| Genspark | Super agent | Full | 30+ (orchestrated) | Web app | Free + credits | Slides, reports, calls |
| Perplexity Computer | Research | High | 19+ (orchestrated) | Perplexity app | $20 / $200/mo | Long research workflows |
| Salesforce Agentforce | Enterprise | High | Atlas + partners | Salesforce | Usage-based | CRM-grounded agents |
| Copilot Studio | Enterprise | High | OpenAI + MS | Microsoft 365 | Consumption | Low-code MS agents |
AI agent frameworks for developers
If you’re building your own agent rather than using one, the choice is a framework or SDK. The field settled around a handful in 2026 (JetBrains, Morph):
- Claude Agent SDK — production agents on the Claude Code harness (Python/TypeScript): tool loop, subagents, sessions, MCP, human-in-the-loop. Now sits under Anthropic Managed Agents, which add scheduling and rubric-based outcome grading.
- OpenAI AgentKit + Agents SDK — AgentKit is the build/deploy/optimise toolkit; the Agents SDK (successor to Swarm) is the code-first path for long-horizon, sandboxed agents. Note: OpenAI’s visual Agent Builder is being wound down, with discontinuation from 30 November 2026 — build in the SDK for longevity.
- LangGraph / LangChain — the most flexible orchestration layer, with deep control over graphs, state and tool routing; the default for complex custom workflows.
- CrewAI — role-based “crews” of cooperating agents; popular for multi-agent task decomposition.
- Google Agent Development Kit (ADK) and Microsoft Agent Framework — the first-party paths for teams standardising on Gemini or Azure/Microsoft stacks respectively.
- n8n — the no-/low-code option for wiring agents into automation pipelines without writing a framework yourself.
The interoperability layer matters as much as the framework: Model Context Protocol (MCP) for tools and the open Agent Client Protocol (ACP) for running one vendor’s agent inside another’s tool are becoming the connective tissue, and are a hedge against being locked to one model provider.
Use-case specific recommendations
For everyday tasks and non-technical users
Winner: ChatGPT Agent (from $20/month)
The most capable general agent in the app most people already use — it browses, books, fills forms and edits files with confirmation checkpoints. Alternatives: Gemini’s agent if you live in Google Workspace; Manus if you want a finished report or deck rather than a chat.
For agentic coding
Winner: Claude Code + Opus 4.8 (from $20/month)
The developer favourite paired with the best available coding model. Alternatives: Grok Build for terminal-native xAI users, OpenAI Codex with GPT-5.5 for CLI/DevOps, Antigravity 2.0 for parallel multi-agent work, and Devin for delegating well-scoped tickets. Full breakdown in our best AI for coding guide.
For autonomous research and deliverables
Winner: Manus, with Genspark close behind
Both take a goal and return a finished artefact — a research report, a slide deck, a populated spreadsheet — by orchestrating multiple models. Manus leads on general task breadth; Genspark is strong on formatted outputs and even automated calls. For sourced research specifically, Perplexity Computer is the pick.
For enterprise process automation
Winner: Salesforce Agentforce (CRM) or Microsoft Copilot Studio (Microsoft 365)
Choose by the system your company already runs on: Agentforce for Salesforce-grounded customer and internal agents; Copilot Studio for low-code agents over Microsoft Graph. Both ground the agent in your own data, which is what actually drives return. See our best AI for business guide.
For developers building custom agents
Winner: Claude Agent SDK or OpenAI Agents SDK, with LangGraph for complex orchestration
Use a first-party SDK when you’re standardising on one model family and want the shortest path to production; reach for LangGraph or CrewAI when you need fine-grained control over multi-agent graphs. Build on MCP so your tools stay portable across models.
For computer-use and desktop automation
Winner: the OSWorld leaders — Claude Opus / GPT-5.5
For agents that must operate real applications, pick a model near the top of OSWorld (Opus-class ~73%, GPT-5.5 ~78% Verified) and expect reliability on routine flows but supervision on edge cases — frontier computer-use only just reaches the human range.
For value and self-hosting
Winner: GLM-5.2 or MiniMax M3 behind an open framework
For data-sovereign or cost-sensitive agents, run an open-weight model built for long-horizon tool use — GLM-5.2 (MIT) or MiniMax M3 — inside LangGraph or the Claude Agent SDK’s open tooling. You trade a few points of ceiling for control and roughly a tenth of the cost.
For the raw capability ceiling
Winner: Claude Fable 5
Now restored to general availability, Fable 5 posts the highest agentic-coding scores of any model. Reserve it for the hardest autonomous work — at twice Opus 4.8’s price and with a tighter safety classifier, it’s a ceiling to reach for, not an everyday default.
Pricing comparison: what you’ll actually pay
Agent tools (typical monthly cost)
| Tool | Free tier | Paid | Billing model |
|---|---|---|---|
| ChatGPT Agent | No (Plus required) | $20 / $200 | Subscription |
| Claude Code | No | $20–200 (or API) | Subscription + credits |
| Antigravity 2.0 | Yes | $20 / $100 | Subscription |
| Grok Build | No (SuperGrok) | $30 (beta) | Subscription |
| Manus | Yes (300 daily credits) | $20 / $40 / $200 | Credits |
| Devin | No | From $20 (usage) | Compute units (~$2.25) |
| Genspark | Yes | Credit-based | Credits |
| Perplexity Computer | No | $20 / $200 | Subscription |
| Salesforce Agentforce | No | Enterprise | Usage-based |
| Copilot Studio | With M365 | Consumption | Message consumption |
Models behind agents (per million tokens, USD)
| Model | Input | Output | Notes |
|---|---|---|---|
| MiniMax M3 | $0.30 | $1.20 | Open weights, strong terminal/tool agent |
| Gemini 3.1 Pro | $2.00 | $12.00 | Cheapest frontier-adjacent |
| GPT-5.4 | $2.50 | $15.00 | OSWorld 75.0%; standardized SWE-bench Pro #1 |
| Grok 4.3 | $1.25 | $2.50 | Powers Grok Build; cheapest flagship |
| Claude Opus 4.8 | $5.00 | $25.00 | Best available agentic default |
| GPT-5.5 | $5.00 | $30.00 | Tool-use / terminal / τ²-bench leader |
| Claude Fable 5 | $10.00 | $50.00 | Highest agentic ceiling; restored 1 July |
Cost strategy: agents multiply token use — every step, tool call and retry bills. Pin routine steps to a cheap or open model, reserve a flagship for the hard reasoning, and cap agent runs with step limits and human checkpoints. Credit-based tools (Manus, Devin) can surprise you on complex jobs, so test on a small task before scaling.
What users and developers actually think
The reliability gap is the headline
Adoption plans are aggressive, but delivered value lags. Only ~17% of organisations have deployed agents, and Gartner expects 40%+ of agentic projects to be cancelled by 2027 on cost, unclear value and weak controls (Gartner). The winners are teams that aim an agent at a narrow, measurable process rather than “automate everything”.
”Agent” is an overloaded word
With only ~130 of thousands of “agentic” vendors shipping anything genuinely autonomous, buyers have learned to test the claim. The reliable end of the market is narrow and concrete — coding agents, CRM/service agents, research agents — while the “does anything” super-agent category is more impressive in demos than in unattended production.
Benchmarks are under scrutiny
The UC Berkeley reward-hacking result (April 2026) hardened a widespread developer instinct: don’t trust agent leaderboards. Practitioners increasingly run small private evals on their own tasks and weight real behaviour — does the agent recover from errors, does it know when to stop, does it flag uncertainty — over headline scores.
Coding agents are where trust is highest
The clearest ROI stories are in software. Devin’s 8–12x migration gains and Claude Code’s developer-favourite status show the pattern: agents work best where tasks are well-specified, verifiable (tests pass or they don’t), and bounded. Independent reviews still land even the best autonomous engineers around 7.5/10, with human oversight assumed.
Recent launches reshaping the market (Apr–Jul 2026)
Grok Build enters the coding-agent race. xAI shipped a terminal-native coding CLI (grok-build-0.1, 256K context, up to eight parallel subagents), making it a three-way contest with Claude Code and Codex (CIO Dive).
Fable 5 restored to general availability (1 Jul). Anthropic’s frontier agentic model returned after its 12–30 June export-control suspension was lifted; Mythos 5 is cleared for approved US organisations but stays trusted-access only (Anthropic).
GPT-5.6 Sol previews (26 Jun). OpenAI’s most capable model tops its own Terminal-Bench harness but remains a government-coordinated limited preview — vetted partners only, no ChatGPT access yet.
Claude Agent SDK credits go live (15 Jun). Anthropic split programmatic agent usage onto a separate credit pool billed at API rates, formalising the SDK as a first-class product (Morph).
Project Mariner shut down (4 May), folded into Gemini. Google retired its standalone browser agent and moved its computer-use capabilities into Gemini Agent and Antigravity 2.0 (Google).
Antigravity 2.0 launches at I/O (19 May). A standalone agent-first platform — desktop, CLI, SDK and managed execution — with Managed Agents in the Gemini API spinning up a tool-using agent in one call.
OpenAI AgentKit ships; Agent Builder to wind down. The build/deploy toolkit launched, but OpenAI is discontinuing the visual Agent Builder from 30 November 2026, steering developers to the code-first Agents SDK.
Agent benchmarks shown to be gameable (12 Apr). UC Berkeley’s RDI demonstrated reward-hacking across all eight major agent benchmarks — the reliability wake-up call of the year (benchmarkingagents.com).
Frequently asked questions
What’s the best AI agent right now?
It depends on the job. For general online tasks, ChatGPT Agent (in ChatGPT Plus, $20/month) is the most capable all-rounder. For coding, Claude Code with Opus 4.8 is the developer favourite. For autonomous deliverables like reports and decks, Manus leads. For enterprise process automation, Salesforce Agentforce or Microsoft Copilot Studio, chosen by the software you already run. There is no single winner — match the agent to the task.
What is an AI agent, and how is it different from a chatbot?
A chatbot answers; an agent acts. An AI agent is a model given tools (a browser, a code sandbox, APIs), memory and a goal, and allowed to take multiple steps on its own until the task is done — browsing sites, running code, editing files, calling services. A chatbot responds to one prompt at a time; an agent plans and executes a multi-step workflow with limited supervision.
What’s the best AI agent for coding?
Claude Code paired with Claude Opus 4.8 — the most-loved tool plus the best available model. Close alternatives are OpenAI Codex with GPT-5.5, xAI’s Grok Build, Google Antigravity 2.0 for parallel agents, and Devin for fully unattended, ticket-scoped work. See our best AI for coding guide for the full comparison.
What’s the best free AI agent?
Manus has the most useful free tier among autonomous agents — 300 daily credits on Manus 1.6 Lite. Google Antigravity 2.0 has a free tier for building and running agents, and Gemini’s agent features are available on its generous free plan. For self-hosting at zero licence cost, run an open-weight model like GLM-5.2 (MIT) behind an open framework.
Are AI agents reliable enough to trust unsupervised?
Not fully. Frontier agents now reach human range on routine computer-use tasks and resolve well-specified coding tickets, but they still fail on edge cases, novel problems and ambiguous goals — and Gartner expects 40%+ of agentic projects to be cancelled by 2027. Treat agents as capable but supervised: aim them at bounded, verifiable tasks, keep human checkpoints on anything consequential, and run your own eval before trusting a benchmark.
What’s the best AI agent for business?
Match it to your stack. Salesforce Agentforce is the most-deployed enterprise platform for CRM-grounded customer and internal agents; Microsoft Copilot Studio is the low-code choice for the Microsoft 365 majority. Both ground agents in your own data, which is what drives measurable return. Full detail in our best AI for business guide.
Which AI model is best for building agents?
For tool-use and terminal work, GPT-5.5 leads (78.2% Terminal-Bench 2.1, 98% τ²-bench Telecom). For long-horizon and coding agents, Opus 4.8 is the pragmatic best and Fable 5 the ceiling. For value or self-hosting, MiniMax M3 and GLM-5.2 bring real agentic capability at open-weight prices.
Is Grok Build better than Claude Code?
Not yet, on current evidence. Grok Build is a capable, terminal-native newcomer with a strong parallel-subagent design, but it’s beta-stage with a smaller ecosystem. Claude Code has the developer-favourite track record and access to Opus 4.8 and the restored Fable 5. Grok Build is worth watching, especially for xAI-ecosystem developers; Claude Code is the safer pick today.
How much do AI agents cost?
Consumer and coding agents start around $20/month (ChatGPT Agent, Claude Code, Manus Standard). Autonomous engineers bill by usage — Devin’s compute units run ~$2.25 each, and heavy runs add up fast. Enterprise platforms (Agentforce, Copilot Studio) are usage- or consumption-based. Because agents multiply token and tool use, the real cost driver is how many steps you let them take — cap runs and route routine steps to cheap models.
Can AI agents replace employees?
Not wholesale. Agents now handle a growing slice of well-specified, verifiable work — migrations, ticketed bug fixes, customer-service flows, research compilation — and are reshaping roles toward specifying, reviewing and orchestrating rather than doing every step. But they struggle with novel judgement, ambiguity and accountability, trust remains low, and most autonomous deployments still assume human oversight. The near-term reality is augmentation of well-scoped tasks, not replacement of whole roles.
The future: what’s coming in late 2026
Autonomy becomes the default interface. ChatGPT Agent, Antigravity, Grok Build and Devin point to a shift from prompting to delegating — you review outcomes, not keystrokes.
Better evaluation becomes the bottleneck. The reward-hacking finding makes trustworthy agent benchmarks a research priority; expect private, task-specific evals and outcome-graded harnesses to matter more than public leaderboards.
Model access turns strategic. With the Fable 5 suspension and GPT-5.6 Sol’s limited preview showing that who’s allowed to use a model now shapes the toolchain, interoperability layers — MCP for tools, ACP for cross-vendor agents — become a hedge.
The agent-washing shakeout arrives. As buyers demand measurable outcomes and 40% of projects get cut, the market should consolidate around the concrete, verifiable categories — coding, service, research — and away from the everything-agent hype.
Conclusion: how to choose in July 2026
The agent market is real, useful and badly over-marketed at the same time. Pick by task, insist on measurable outcomes, and supervise.
- Best general agent: ChatGPT Agent — most capable all-rounder, in the app you already use.
- Best for coding: Claude Code + Opus 4.8; watch Grok Build and Antigravity 2.0.
- Best autonomous “super agent”: Manus, with Genspark close for finished deliverables.
- Best for enterprise: Salesforce Agentforce or Microsoft Copilot Studio, by your stack.
- Best for builders: Claude Agent SDK or OpenAI Agents SDK, LangGraph for complex orchestration.
- Best underlying model: Opus 4.8 for everyday, GPT-5.5 for tool-use, Fable 5 for the ceiling.
- Best value / self-host: MiniMax M3 or GLM-5.2 behind an open framework.
The tools genuinely work now for bounded, verifiable tasks — and genuinely don’t for the “do everything unattended” pitch. The variable that decides your result isn’t the logo on the agent; it’s whether you point it at a measurable job, keep a human in the loop, and test it on your own work before you trust a benchmark. For the models and coding tools behind these agents, see our best AI models and best AI for coding rankings.
This guide is updated as agent tools, models and benchmarks evolve. Agent benchmark scores are unusually unreliable — vendor numbers run above standardized harnesses and researchers have shown the major benchmarks can be gamed — so we cite sources and weight real-world behaviour over leaderboards. Pricing, availability and model access change frequently; verify with the provider before purchasing.