AI Coding Tools Hub: Assistants, IDEs and Coding Agents Side by Side

AI coding tools now span editor assistants, agentic IDEs, terminal agents, cloud agents and prompt-to-app builders, and the market changed shape several times during 2026. This hub lists the major tools from official vendor pages, with the same fields for every product: owner, category, surfaces, entry price, team or enterprise route and current status. It also records the 2026 transitions that affect buying decisions, such as the Gemini Code Assist sunset and the Amazon Q Developer end of support. It is written for engineering leaders, platform teams and developers who need a factual starting point before a trial or a procurement review. Tools are listed alphabetically; order is never a ranking.

Last checked: 2026-11-15.

This is an independent reference page from NHI Mgmt Group based on public provider documentation; it is not sponsored or endorsed by any provider.

Key takeaways

• Five categories, overlapping fast. Most vendors now ship more than one surface: an editor or IDE, a terminal agent, a cloud agent that works on repositories, and admin controls for teams.
• Usage, not seats, is the real unit. Nearly every plan meters model use through credits, tokens, rate windows or a dollar allowance on top of the seat price, so the headline price is only part of the cost.
• 2026 brought hard deadlines. Google ended sale of Gemini Code Assist Standard and Enterprise on 9 October 2026 and points customers to Antigravity; AWS set 30 April 2027 as the end of support for Amazon Q Developer IDE plugins and paid subscriptions, with Kiro named as the migration path.
• Bundles are changing what a tool is. Cursor is sold as part of The X Subscription alongside Grok and X, Claude Code is included in Anthropic's paid Claude plans, and Codex is included in ChatGPT plans.
• Security controls sit with the customer. SSO, SCIM, audit logs, training opt-outs and spend caps are plan dependent; confirm them per plan and read the tool's NHI Mgmt Group security guide before rollout.

What counts as an AI coding tool?

An AI coding tool is software that uses a large language model to write, edit, explain, review or run code on a developer's behalf. The category now covers five product shapes, and one vendor may sit in several.

The distinction matters for governance. An editor assistant suggests; an agent acts, with the developer's or a service account's permissions. The AI Coding Assistants vs Autonomous Coding Agents guide covers that difference in depth, and Vibe Coding Explained covers the app builder style of work.

Which AI coding tools exist in 2026?

The table uses identical fields for every tool. Entry price is the lowest paid individual plan on the vendor's official pricing page in US dollars per month, excluding tax. As of 15 November 2026.

Full plan detail, usage units and overage rules for each tool are on the AI Coding Tools Pricing Tracker.

What does each tool offer?

Each entry uses the same four points: surfaces, how usage is metered, enterprise controls stated on the official page, and the NHI Mgmt Group security guide where one exists.

Antigravity (Google)

• Surfaces: Antigravity IDE, Antigravity CLI, and Antigravity IDE extensions for VS Code and JetBrains, according to Google's Code Assist sunset page.
• Usage: weekly rate limits, unlimited Tab completions and Command requests on the free tier; Google AI Pro and Ultra add higher limits and a flexible AI credit pool.
• Enterprise: spend caps and centralised administrative controls with eligible plans through Gemini Enterprise.
• Security guide: the Gemini CLI and Gemini Code Assist Security Guide covers the Google coding estate.

Bolt (StackBlitz)

• Surfaces: browser based builder with website hosting, databases and custom domains on paid plans.
• Usage: monthly token allowance; paid plans keep unused tokens for up to two months while the subscription is active.
• Enterprise: SSO, audit logs, compliance support, data governance and retention policies on Enterprise.
• Security guide: the AI App Builder Security Guide.

Claude Code (Anthropic)

• Surfaces: CLI and VS Code extension; also available through Amazon Bedrock, Google Cloud's Agent Platform and Microsoft Foundry, billed per token to the cloud account.
• Usage: on subscription plans, rolling five-hour and weekly windows shared with Claude chat and Cowork; on the Claude Console, per-token API billing. Anthropic's cost documentation states an enterprise average of around $13 per developer per active day and $150 to $250 per developer per month.
• Enterprise: spend limits at organisation, group and member level, a spend report, the Enterprise Analytics API, and managed settings.
• Security guide: the Claude Code Security Guide.

Cursor

• Surfaces: the Cursor editor, cloud agents and automations, and AI code reviews on Premium and above.
• Usage: a usage pool shared across Cursor and Grok that refreshes weekly; third-party models are billed at API list price from prepaid credit. Cursor's documentation states that Cursor is sold as part of The X Subscription, "one subscription for Grok, Cursor, and X".
• Enterprise: invoice billing, SCIM seat management, spend controls, audit logs, and repository, model and MCP access controls.
• Security guide: the Cursor Security Guide.

Devin (Cognition)

• Surfaces: Devin Cloud and Devin Desktop; Devin's documentation describes Devin Local as the primary local agent for Devin Desktop, sharing its harness with Devin CLI, with enterprise administration through the Windsurf dashboard.
• Usage: plan quotas, with extra usage on paid plans consumed at API pricing; Pro and Max can use a ChatGPT plan.
• Enterprise: SCIM, role-based access control, admin controls and a dedicated deployment option including VPC.
• Security guides: the Devin Security and Windsurf Security Guide pages.

GitHub Copilot

• Surfaces: IDE completions and chat, agent mode, CLI, Copilot cloud agent, code review, and delegation to third-party agents (Claude Code and Codex) on Pro and above.
• Usage: GitHub AI Credits at $0.01 each, split into a base and a flex allotment; Business and Enterprise credits are pooled.
• Enterprise: Business and Enterprise organisation plans; audit logs are listed from Pro+.
• Security guide: the GitHub Copilot Security Guide.

JetBrains AI

• Surfaces: AI features inside JetBrains IDEs from version 2025.1; JetBrains states that Junie is now part of JetBrains Air.
• Usage: AI credits worth $1 each, a monthly quota per tier, and top-up credits valid for 12 months; credits billed at public API rates are also offered.
• Enterprise: AI Enterprise through JetBrains IDE Services, with a shared credit pool and admin monthly limits.
• Security guide: the JetBrains AI Assistant and Junie Security page.

Kiro (AWS)

• Surfaces: agentic IDE and CLI for spec-driven development; AWS lists agentic coding, inline chat, terminal integration and MCP support.
• Usage: monthly credits per plan, with add-on credits at $0.04.
• Enterprise: credit pooling for teams of two or more.
• Security guide: the Amazon Q Developer and Kiro Security Guide.

Lovable

• Surfaces: browser builder with Git sync, publishing, custom domains and the Lovable API on Business.
• Usage: monthly plan credits that roll over while active and expire two months after issue, plus daily build credits and monthly Cloud, AI and connector grants that do not roll over.
• Enterprise: SSO and role-based access on Business; SCIM, audit logs, scheduled security scans and EU inference on Enterprise.
• Security guide: the Lovable Security Guide.

OpenAI Codex

• Surfaces: Codex on the web, CLI, IDE extension, desktop app and iOS, included across ChatGPT plans.
• Usage: estimated local messages per five-hour window that vary by model, weekly limits, and credits; API-key use in the CLI, SDK or IDE extension is billed at API pricing, without cloud features such as GitHub code review.
• Enterprise: Enterprise and Edu with flexible pricing scale on workspace credits without fixed rate limits.
• Security guide: the OpenAI Codex Security Guide.

Replit

• Surfaces: browser workspace with Replit Agent, hosting and databases, with database rollback up to 28 days on Pro.
• Usage: a monthly dollar allowance toward more capable models, Free Mode chat and project limits, and effort-based pay-as-you-go pricing.
• Enterprise: SSO, SCIM, audit logs, privacy controls, single-tenant and private networking options.
• Security guide: the Replit Security Guide.

v0 (Vercel)

• Surfaces: browser builder with deployment to Vercel, Design Mode visual editing and GitHub sync.
• Usage: monthly credits per user plus $2 of free credit on daily login on paid plans; extra credits are shared across the team.
• Enterprise: training opt-out by default on Business; SAML SSO, role-based access control and no training on customer data on Enterprise.
• Security guide: the AI App Builder Security Guide.

What changed in 2026?

Dated changes stated on official pages, ordered by date. As of 15 November 2026.

Release history for each product is covered on the AI Coding Tools Timeline. Teams retiring Amazon Q Developer plugins should also finish in-progress Java and .NET transformations before 30 April 2027; AWS points to AWS Transform for that work.

How do I choose an AI coding tool?

These are factual criteria, not a recommendation. Score each candidate against the ones that matter to your organisation.

• Small teams: individual or small team plans with a pooled allowance usually suffice; watch for features such as SSO that appear only on business tiers.
• Large organisations: prioritise SCIM, audit logs, spend caps per user or group, and analytics APIs, and pilot with a small group to set a usage baseline before rollout, as Anthropic's cost documentation recommends for Claude Code.
• Regulated environments: check training defaults, data residency options and dedicated or VPC deployment, which several vendors offer only on enterprise plans.

The AI Coding Assistant Buyer's Guide turns these criteria into evaluation questions, and the AI coding assistant rollout playbook covers a phased rollout.

Where identity and security fit

Coding agents hold credentials: personal access tokens, API keys, cloud roles and OAuth grants to repositories. Every tool on this page is a non-human identity footprint inside the software delivery pipeline.

• Token and key sprawl: API keys for "bring your own key" modes and CI use end up in shell profiles and configuration files. See AI Coding Tools and Secrets Sprawl.
• Configuration as an attack surface: agent rule files and MCP configuration change agent behaviour. See Coding Agent Configuration File Attacks.
• Shadow adoption: free tiers make unsanctioned use easy. See Unsanctioned AI Coding Assistants.
• Agent controls: the AI IDE and Coding Agent Security Guide covers sandboxing, approvals and logging across tools.

Practitioner checklist

• Build an inventory of every AI coding tool in use, including free tiers, CLIs and browser builders, with the owner of each.
• Confirm the plan level that provides SSO, SCIM and audit logs for each sanctioned tool, and move business use onto it.
• Record the usage unit and limit for each plan, and set spend caps per user or group before broad rollout.
• Check each tool's training and retention defaults on your plan and turn on opt-outs where offered.
• Track announced lifecycle dates (Gemini Code Assist, Amazon Q Developer) and set a migration owner and deadline.
• Issue agent credentials with least privilege and short lifetimes, and keep API keys out of repositories and dotfiles.
• Review agent rule files and MCP server configuration as code, through pull requests.
• Re-check official pricing pages before each renewal; prices on every page here are stated as subject to change.

Standards and references

All sources accessed 15 November 2026.

• AWS: Amazon Q Developer end of support announcement
• Google Cloud: Gemini Code Assist Standard and Enterprise sunset
• Google for Developers: Gemini Code Assist for individuals deprecation
• Google: Antigravity pricing
• StackBlitz: Bolt pricing
• Anthropic: Claude pricing
• Anthropic: Claude Code, manage costs effectively
• Cursor: Models and pricing documentation
• Cognition: Devin pricing
• Cognition: Devin Local documentation
• GitHub Docs: Plans for GitHub Copilot
• JetBrains: AI Assistant licensing and subscriptions
• AWS: Kiro pricing
• Lovable: Subscription plans
• OpenAI: ChatGPT and Codex pricing
• Replit: Pricing
• Vercel: v0 pricing

Related NHI Mgmt Group resources: AI Coding Tools Pricing Tracker · AI Coding Tools Timeline · Coding LLMs Compared · Vibe Coding Tools Compared · AI Coding Agents Security Guide