This week's changes

Week of November 2, 2026

No ranking changes this week. Compared with October 26, 2026.

Notes

• No rank moves, new entries, drops or pick changes this week.
• Security: on Oct 29, 2026, tampered copies of Anthropic's Claude Code extension (2.1.312 and 2.1.313) were served from Cursor's VS Code extension marketplace for about 11 hours (2:27am to 1:35pm PT) to macOS and Windows users. The fake build tried to plant an SSH backdoor. Anthropic didn't build it, and the official VS Code Marketplace and npm packages weren't affected. Cursor pulled it, rebuilt the marketplace and published a check script. If you installed or updated that extension in Cursor that day, follow Cursor's steps: uninstalling alone doesn't remove it. We're keeping Cursor's rank: this is a distribution incident, not a change in the editor itself. Sources: cursor.com/blog/claude-code-extension-security-notice and microsoft.com/en-us/security/blog/2026/10/30/inside-malicious-claude-code-branded-cursor-extension-persistent-ssh-backdoor/
• Bolt.new pricing is now verified: Free has 1M tokens a month (300K a day), Pro is $25/mo with 10M tokens that roll over, Teams is $30 per member per month. Source: bolt.new/pricing
• OpenAI Codex and ChatGPT: starting in December, Plus, Pro and Business plans will share one token-based allowance across ChatGPT and Codex (Instant chat excluded). OpenAI hasn't published the allowance sizes. Source: help.openai.com/en/articles/20001625-understanding-chatgpt-and-codex-usage-limits
• ChatGPT Pro ($100, $200 and $500 tiers) and Business now include monthly OpenAI API credits: $100 / $200 / $500 for Pro, and $20 per Standard or $100 per Premium Business seat, capped at $500 per workspace. The credits don't cover Codex through API keys. Source: help.openai.com/en/articles/20001626-api-credits-benefit-for-eligible-chatgpt-subscriptions
• GitHub Copilot: GPT-6.1 Terra became generally available for Pro, Pro+, Max, Business and Enterprise on Oct 26. Source: github.blog/changelog/2026-10-26-gpt-6-1-terra-in-github-copilot/
• Claude Code 2.1.311 to 2.1.316: a setting that blocks writes outside your project folder, and fast mode for Opus 4.8 is deprecated (removed Nov 30, 2026; requests will error, not fall back). Anthropic says to move to Opus 5.5. Source: Claude Code changelog, via aicoding-guide.com/en/posts/weekly-2026-11-01/
• OpenAI Codex CLI 0.165 and 0.166: a built-in Google Cloud Vertex AI provider, memories on by default for new profiles, and codex queue --resume. No price change. Source: OpenAI Codex changelog, via aicoding-guide.com/en/posts/weekly-2026-11-01/
• Checked: Lovable's pricing page still doesn't show dollar prices in the text we can read, so Lovable's paid prices stay unverified. No scores changed this week.

Week of October 26, 2026

No ranking changes this week. Compared with October 19, 2026.

Notes

• No rank moves, new entries, drops or pick changes this week. We re-checked prices and plans across all 8 categories.
• Correction: Google Antigravity. Gemini 4 Argon is available to Google AI Pro and Ultra subscribers in the Antigravity desktop app and CLI, not the standalone Antigravity IDE. We previously said Ultra only and didn't note the IDE limit. Source: antigravity.google/blog/gemini-4-argon-in-google-antigravity/
• GitHub Copilot Free: GitHub's FAQ now describes Free as 2,000 completions and access to entry-level models per month, with limited chat and agent use. The old 50-chat-requests figure is gone, so we removed it. Paid prices unchanged (Pro $10, Pro+ $39, Max $100). Source: github.com/features/copilot/plans
• Lovable Free: the pricing page now also lists 4 free credits a month for AI features built into your app, on top of 5 daily build credits and 20 Cloud credits. Source: lovable.dev/pricing
• Claude Code shipped releases 2.1.303 to 2.1.310 (Oct 18 to 23): per-repo Quick/Standard/Deep code review levels, a safety-event telemetry signal, and sandbox and permission hardening. No price change. Source: Claude Code changelog, via javarubberduck.com/ai/news-2026-10-25-dev-tools/
• OpenAI Codex CLI 0.164 (Oct 23) added saved prompt presets, enterprise MCP sign-in and wider credential masking. No price change. Source: OpenAI Codex changelog, via javarubberduck.com/ai/news-2026-10-25-dev-tools/
• Checked and unchanged: Cursor, Claude, GitHub Copilot paid plans, v0 and Lovable paid plans. No scores changed this week.

Week of October 19, 2026

0 rank moves, 1 new, 0 dropped, 0 tier changes. Compared with October 15, 2026.

New entries

• Claude Code Review (Code review & testing, #4, B): Generally available Oct 15, 2026; multi-agent reviews with a verification step, Quick/Standard/Deep levels

Notes

• New entry: Claude Code Review (Code review & testing, #4, B). Anthropic made it generally available for Claude Team and Enterprise on Oct 15, 2026. It's billed per review on usage credits, so we flag its pricing. Source: claude.com/resources/articles/code-review-is-now-generally-available
• Gemini API: Gemini 4 Argon went GA on Oct 13, 2026 at an introductory $2 in / $10 out per million tokens through Nov 30, then $4 / $20. It isn't on the free tier. Pricing on our board is now verified. Source: ai.google.dev/gemini-api/docs/pricing
• GitHub Copilot: Gemini 4 Argon is GA in Copilot for Pro+, Max, Business and Enterprise (Oct 13). GitHub will retire some older models, including Claude Opus 4.8, Claude Opus 5 and GPT-5.6 Terra, on Nov 16, 2026. Source: github.blog/changelog/month/10-2026/
• OpenCode: version 2.0 became the main release on Oct 12, 2026. Plugins written for 1.x don't work on 2.0. Source: thenewstack.io/opencode-2-shared-service/
• Correction: Windsurf was renamed Devin Desktop by Cognition on June 2, 2026, before our first ranking. Same IDE; Cognition says plans and pricing didn't change. We now link the Devin Desktop FAQ. Source: docs.devin.ai/desktop/devin-desktop-faq
• New section: AI tools for small metal shops (welding, fabrication, misc/structural steel, small machine shops) at gettooltier.com/shop. 10 tools in 6 categories, 4 answer pages, and a Shop AI Stack Audit. Prices checked on the vendors' pages Oct 19, 2026. Hands-on tests by our founder (a working structural steel fitter) have not happened yet; each tool says so. The coder rankings above are unchanged.
• Checked and unchanged: Cursor, Claude, GitHub Copilot, CodeRabbit, v0, Lovable, Granola and ElevenLabs pricing pages. No scores changed this week.

Week of October 15, 2026

Baseline week: 29 tools across 8 categories published. Nothing to compare against yet.

Where things stand

• AI coding editors/IDEs: Cursor is our pick
• Coding agents (CLI/cloud): Claude Code is our pick
• App builders (prompt-to-app): Lovable is our pick
• AI model APIs: Anthropic Claude API is our pick
• Code review & testing: CodeRabbit is our pick
• Design-to-code / UI generation: v0 by Vercel is our pick
• Docs & research for devs: Gemini Deep Research + Notebook is our pick
• Dev productivity: Granola is our pick

18 of 29 entries have at least one pricing or plan detail we couldn't fully confirm. Those are flagged with ⚠ on the cards.

Notes

• First published ranking (v1). This is the baseline: future weeks are compared against it.
• Scores are editorial judgments informed by public benchmarks, October 2026 reviews, and official pricing pages. They are not hands-on lab tests yet.