Every week we publish what changed across the 1,695 AI tools we have scored, plus a handful worth your attention. This is the issue for Sep 29 - Oct 5, 2026, sent to subscribers and archived here in full.
What we published
Best MCP Servers for Web Scraping in 2026: Bright Data vs Firecrawl vs Apify vs Browserbase, Priced per 1,000 Pages for AI Agent Teams
For a team feeding live web pages to AI agents, 100,000 pages a month costs $99 on Firecrawl Standard, $142.50 on Bright Data pay as you go, at least $150 on Apify's default fetch tool, and $382.33 on Browserbase's browser-hour MCP. Below 71,000 pages, Bright Data costs less. Every price read from the vendor's own page on 27 September 2026.
Best MCP Servers for Claude Code and Cursor Teams in 2026: 12 Maintained Picks, 6 With No Price Tag, and What the Other 6 Cost a Month
12 maintained MCP servers for developers and small teams on Claude Code, Cursor or another coding agent, each checked against its own repository, package registry and pricing page on 27 September 2026. 6 carry no price of their own. The other 6 start free, and at entry paid tiers the 5 with a monthly price cost one developer $86 a month. 3 of the 6 most-installed entries in the open Claude Code Templates registry call npm packages that are deprecated or missing.
Best Claude Code Subagents in 2026: The 12 Most Installed, Which Ones Save an Engineering Team Review Time, and What Each Costs to Run
The 12 most installed Claude Code subagents, from the open Claude Code Templates registry (aitmpl.com), 2026-09-26, audited file by file on 27 September 2026. 10 of 12 can edit your code and 9 run on whatever model the developer's session uses, $2 to $4 per million input tokens on the Anthropic API. The ones that save an engineering team review time are code-reviewer, debugger and test-engineer; the builders need editing first, and context-manager cannot do its job as shipped.
New to the catalogue
- Heretic (LLM Infrastructure, scored 6.8)
- text-to-cad (AI Agents, scored 7.2)
- Strix (AI Agents, scored 7.3)
- Pascal Editor (AI Agents, scored 7.4)
- Beads (AI Agents, scored 6.9)
- SGLang (LLM Infrastructure, scored 7.2)
- PageIndex (RAG & Search, scored 7.2)
- tuios (AI Agents, scored 6.7)
Five worth your time
| Tool | Category | Score |
|---|---|---|
| Taskade | AI Agents | 8.5/10 |
| Chat Summary for LiveChat | AI Assistant | 8.3/10 |
| Exemplary AI | AI Video | 8.3/10 |
| Mails.ai | AI Coding | 8.3/10 |
| Speak Ai | Productivity | 8.3/10 |
Taskade. One prompt to a live app - CRMs, dashboards, client portals run by AI agents and workflow automations. Build free with Taskade Genesis, no credit card.
Chat Summary for LiveChat. LiveChat® is the best AI live chat software for business, perfect for ecommerce and B2B. It powers real-time website support, boosting sales and customer satisfaction.
Exemplary AI. Turn podcasts, webinars, and videos into shareable clips, transcripts, summaries, and social posts. Exemplary AI does it all.
Mails.ai. Self-serve AI email API + MCP for agents. Per-agent addresses, per-agent reputation, native prompt-injection scanning on every inbound.
Speak Ai. Capture every call, meeting, and role play. Score it on your methodology. Turn it into action. Free 7-day trial. No credit card required.
The honest limit
We have scored 1,695 tools across 24 categories, and we have not tested every one to the same depth. Where a score leans on vendor documentation rather than hands-on use, the review says so on the page.
Want this in your inbox each week? Create a free account and it comes with it. We sell no rankings, and every affiliate link is labelled.
Comments
Sign in to join the discussion.
Login to comment