Gemini climbs to 13/36 as Copilot slips to 18/36
Gemini moved up 1 point to 13/36 on the back of a faster coding run, while GitHub Copilot fell 2 points to 18/36 as GitHub’s own notes point to strain in its agent stack. New listings also matter: BrowserAct lands at 14/36 with a managed browser, proxy, fingerprint, and human hand-off layer for multi-step web work. In the same week, Claude Code is up +69.6%, which says operators are still paying attention to agentic coding, but they want systems that can survive real web friction.
Foundation-model news
|
GPT-5.4 · OpenAI
Model release
Researchers at the University of Illinois Urbana-Champaign, UC Berkeley, and Chroma unveiled Harness-1, a 20-billion-parameter open-source search agent built on OpenAI's gpt-oss-20B model. The system is designed to rethink how agents execute complex retrieval tasks. It scores 73% on recalling relevant information correctly, topping GPT-5.4 in the reported test. For those tracking retrieval and agent tooling, this points to open-source systems closing the gap on high-value search workflows.
Source: VentureBeat — https://venturebeat.com/orchestration/researchers-trained-an-open-source-ai-search-agent-harness-1-that-outperforms-gpt-5-4-on-recalling-relevant-information.
|
|
Gemini 3.5 · Google DeepMind
Model release
Google unveiled Gemini 3.5 and a broader set of AI and hardware updates at I/O 2026 in Mountain View on May 19 and 20. The company says the new model pushes harder on reasoning and action-taking as it frames a new Agentic Gemini era. For those tracking agentic AI, this signals where Google wants to place its bets on models that do more than answer questions. The catch is that the summary points to a heavily promotional announcement, so the technical depth behind the branding still needs a closer look.
Source: blockchain.news — https://blockchain.news/news/google-gemini-3-5-ai-updates-io-2026.
|
What moved today
|
13/36 · Adaptive Collaborator
Score re-evaluated after: Gemini 3.5 Flash outruns OpenAI's GPT-5.5 in coding speed, but its mistakes pile up fast.
Trigger: PCMag Australia — "Gemini 3.5 Flash outruns OpenAI's GPT-5.5 in coding speed, but its mistakes pile up fast.".
|
|
18/36 · Adaptive Collaborator
Score re-evaluated after: GitHub's Kyle Daigle says the Copilot boom is straining agents, and GitHub now has a plan.
Trigger: Latent Space (swyx + Alessio Fanelli) — "GitHub's Kyle Daigle says the Copilot boom is straining agents, and GitHub now has a plan.".
|
What was reported, but the rubric didn't move
|
6/36 · Guided Assistant
Mentioned in coverage but score held. Press without verifiable capability change does not move the rubric.
Source: mint — "Why the market isn't buying Salesforce's AI story | Company Business News".
|
|
13/36 · Adaptive Collaborator
IBM and Google Cloud announced a new customer product on June 4 for enterprise buyers. The move follows years of overlap in the enterprise AI market, where each company brings different strengths to the table. For those tracking agentic AI product launches, this matters because the market is shifting toward packaged systems that can be deployed inside real companies. The key question is whether the two companies can finally make AI work at scale for regulated customers.
Source: Belleville News-Democrat — "IBM and Google Cloud say their June 4 launch targets enterprise AI buyers still stuck between models and implementation.".
|
|
7/36 · Guided Assistant
Canva said its new Perplexity Computer integration turns AI research, strategy docs, competitive analysis, and business briefs into editable visual assets. The push comes as creators want faster paths from first draft to finished presentation or campaign. For those tracking AI writing and content tools, this signals a stronger move from text generation into publishing-ready workflows. The big draw is the handoff from research to design without starting over in another app.
Source: MobiGyaan — "Canva integrates with Perplexity Computer to turn AI research into editable campaigns, shrinking the gap to publishable content.".
|
What's new in the directory
|
14/36 · Adaptive Collaborator
BrowserAct provides a managed browser environment for AI agents to browse the web, extract data, and handle blocked pages. It also offers proxy, fingerprint, and human hand-off infrastructure for multi-step web workflows.
Source: published listing, rubric-scored on ingest.
|
|
12/36 · Guided Assistant
HouseWhisper follows up with real estate leads, nurtures them over time, and routes ready prospects to the right agent. It connects to existing CRMs, dialers, and calendars so teams can work from their current stack.
Source: published listing, rubric-scored on ingest.
|
|
9/36 · Guided Assistant
Ylopo is a real estate marketing platform that helps you generate, nurture, and convert buyer and seller leads. It combines custom websites, ads, CRM integration, and AI text/voice follow-up to move leads toward appointments.
Source: published listing, rubric-scored on ingest.
|
From Agentic-News.ai
Agentic-AI industry signal we're tracking — launches, releases, and research from across the ecosystem. The directory rubric hasn't scored these yet; awareness now, evaluation next pass.
|
Macrobond
product launch
Macrobond on June 10 announced general availability of its AI Data Feed Service for financial institutions. The product includes MCP for conversational AI and Skill for programmatic access to licensed macroeconomic data. For those tracking agentic AI product launches, this matters because it puts governed data directly inside AI research and strategy tools. The key wrinkle is that the company is packaging trust and access together, not just raw data.
Via Agentic-News.ai — FinanzNachrichten.de. Read: https://www.finanznachrichten.de/nachrichten-2026-06/68730495-macrobond-launches-ai-data-feed-with-mcp-and-skill-bringing-governed-macroeconomic-intelligence-to-enterprise-ai-workflows-004.htm
|
|
Volante Technologies
product launch
Volante Technologies launched Vol360i Agentic AI across its payments platform and PaaS operations on Monday. The company says the upgrade raises straight-through processing above 95%, speeds exception resolution, and helps manage SLA performance. For those tracking agentic AI, this is a concrete example of autonomy moving into regulated finance workflows. The most notable detail is immediate availability, which makes this more than a roadmap promise.
Via Agentic-News.ai — WBOC TV-16. Read: https://www.wboc.com/online_features/press_releases/volante-technologies-launches-vol360i-agentic-ai-at-the-core-of-payments-setting-a-new-global/article_3cf43da0-06a1-58a6-83de-a0b169c6f698.html
|
|
enso
product launch
enso introduced Agentic Growth Hacking on June 9, 2026, framing it as a new discipline for AI agents to execute go-to-market tasks at scale. The company says its agents can work across places, surfaces, and conversations that human teams cannot reach fast enough. For those tracking the AI agent platform ecosystem, this is a sign vendors are moving from generic agents to packaged business functions. The notable angle is enso trying to name and own an entire category, not just sell a tool.
Via Agentic-News.ai — Barchart.com. Read: https://www.barchart.com/story/news/2374550/enso-introduces-agentic-growth-hacking-a-new-category-for-the-ai-era
|
Biggest movers this week
Listings with the largest week-over-week jump in outbound-click interest
- Claude Code · +69.6% week-over-week. Claude Code is Anthropic's agentic coding tool that lives in your terminal. It understands your entire codebase, makes multi-file edits, runs commands, manages git workflows, and uses MCP for tool integration. Built with a Unix philosophy — it reads, plans, edits, and verifies in a loop. The fastest-growing product in the coding agent category.
- Manus · +55% week-over-week. Manus turns plain-English ideas into production-ready websites and internal apps. It’s aimed at teams that want to launch full-stack products without engineering resources.
- Cline · +26.4% week-over-week. Cline is an AI coding assistant for VS Code that can inspect your project, edit files, run terminal commands, and use a browser while asking for permission at each step. It is aimed at developers working on real codebases who want more than code completion.
- Windsurf · +23.9% week-over-week. Windsurf Editor is an AI-powered IDE for developers that blends chat, autocomplete, and agentic code actions into the editor itself. It’s available for Mac, Windows, and Linux, and is aimed at speeding up day-to-day coding work on real codebases.
- Codex CLI · +15.2% week-over-week. Codex CLI is OpenAI’s coding agent that runs on your computer from the terminal. It can also be used in VS Code, Cursor, Windsurf, or as a desktop app.
Biggest declines this week
Listings with the largest week-over-week drop in outbound-click interest
- AgentSwarms · -40% week-over-week. AgentSwarms is a browser-based learning environment for Agentic AI. You can follow guided lessons, run live demos, and build and save agents and swarms without setup or a credit card.
- Wingman by Emergent · -38.8% week-over-week. Wingman is a messaging-first autonomous AI agent by Emergent. It connects to your email, calendar, Slack, CRM, and GitHub, then handles scheduling, research, sales support, and admin tasks through WhatsApp and Telegram conversations.
- Hermes Agent · -35% week-over-week. Hermes Agent is an open-source autonomous agent from Nous Research that runs on your server and keeps context over time. It can work across chat apps, the CLI, and the browser to handle multi-step tasks and scheduled automations.
Agenticness Leaderboard
This week's most-clicked tools, re-ranked by rubric score. The score is the editorial signal; the popularity is what drew the directory's traffic here this week.
- Cline · 19/36 · Domain Specialist · L3
Cline is an AI coding assistant for VS Code that can inspect your project, edit files, run terminal commands, and use a browser while asking for permission at each step. It is aimed at developers working on real codebases who want more than code completion.
- Windsurf · 19/36 · Domain Specialist · L3
Windsurf Editor is an AI-powered IDE for developers that blends chat, autocomplete, and agentic code actions into the editor itself. It’s available for Mac, Windows, and Linux, and is aimed at speeding up day-to-day coding work on real codebases.
Claude Code is Anthropic's agentic coding tool that lives in your terminal. It understands your entire codebase, makes multi-file edits, runs commands, manages git workflows, and uses MCP for tool integration. Built with a Unix philosophy — it reads, plans, edits, and verifies in a loop. The fastest-growing product in the coding agent category.
Wingman is a messaging-first autonomous AI agent by Emergent. It connects to your email, calendar, Slack, CRM, and GitHub, then handles scheduling, research, sales support, and admin tasks through WhatsApp and Telegram conversations.
What operators are searching for this week
A week's demand signal — what operators are typing into the directory's search. Suggestion-chip defaults + off-topic / greetings filtered.
- "Sales"
- "Semgrep"
- "Construction office"
One tool worth testing
BrowserAct — BrowserAct is the one to test if you need an agent that can browse, extract data, and keep moving through blocked pages. It just entered the directory at 14/36, and its managed browser plus proxy, fingerprint, and human hand-off stack maps to the part of web automation that usually breaks first. Evaluate it against your current browser automation in a 30-minute pilot on one workflow that hits logins, bot checks, or dynamic pages.
|