New AI Tools 2026 (June Week 1): 7 Tested, Ranked and Reviewed

This is Week 1 of June 2026 on AI Tools Radar. If you want new ai tools 2026 without another recycled list, start here. We scanned AIxploria last-ai and TAAFT just-released. We ran autocomplete on tool names. We kept launches that fit our lanes: agents, creators or slides, builders. Model-only API changes live on our models hub. They still get a slot here when search demand spikes.

June directories were heavy on presentation AI, UGC video, and short clip editors. We added one agent (Manus) and one builder test (WebZum) so the week is not only slides and TikTok tools.

Week 1 at a glance

---

Changelog

• 2026-06-02: First June radar published. Slot 7 is SlideAI. OminiGate moved to OpenRouter guide mention only. Added Dokie, VmakeAI, Aleph 2.0, and WebZum from directory scans.
• 2026-06-02: Fact-check pass. Manus and Dokie verdicts aligned to full reviews (Watch). VmakeAI URL → vmake.ai. WebZum URL → webzum.com.

Method in one paragraph

We do not list every directory entry. A tool needs search demand or clear buzz plus a testable signup (or a stable public demo). Full reviews publish only when two or more gate criteria pass (see Editorial Policy): hands-on access, pricing visibility, repeatable workflow, and reader questions we can answer with evidence. This week: seven radar entries, two with full reviews linked (Manus, SlideAI), one model tracked on the living comparison page (Claude Opus 4.8), and four tools on Watch until export, billing, or rights checks finish.

---

1. Manus AI

Verdict: Watch (pilot before daily use; see full review)

Manus AI is still the talk of the town for autonomous agents, even when it is not the newest name on AIxploria. The product at manus.im runs tasks in the background. You describe an outcome. The agent plans steps, uses tools and browsing where allowed, and returns files like CSVs, briefs, or tables. It does not stop at chat-only answers. Reviews in 2026 still mention per-task pricing near two dollars. Users also report paywalls, invite queues, and downtime when load is high.

Search signals: People still type manus ai review, manus ai pricing, and manus ai agent. That is why Manus stays on the radar this week even though our headline verdict is Watch. Directory pages may highlight newer video apps. Search demand keeps Manus visible while Meta and China ownership news settles.

Verdict note: Our Manus AI Review (2026) says Watch for most teams until pricing and compliance are confirmed on your account. Move to Use after two end-to-end tasks without billing surprises.

What we tested (summary): We used the same checklist as our full review. We checked signup friction. We ran one research brief and one structured output request. We compared it to ChatGPT agent mode. Manus shows a Computer UI. You can see browsing and tool actions. That helps if you need an audit trail, not a black-box answer.

Best for: Lead lists, market scans with filters, first passes on decks or reports when you supply clean sources, and repeat monitoring jobs that follow a template.

Skip if: You need same-day uptime every time, nuanced brand voice, or paywalled databases without uploading your own exports.

Pricing signal: Treat billing as verify-live. Screenshot your plan page when you pilot. Failed tasks may still use credits. Rules can change.

What we would test next: Enterprise data policy PDF, refund rules on failed tasks, and whether the same weekly report prompt works the same way across ten runs.

• Lane: Agents
• Competes with: ChatGPT deep research, Perplexity-style research, OpenAI operator features
• Full review: Manus AI Review (2026)

---

2. Dokie AI

Verdict: Watch (presentation lane; full review published)

Dokie AI is an AI presentation maker on AIxploria with strong community votes in late May and early June 2026. The pitch is simple. Turn plain text, notes, or documents into a business-ready PowerPoint deck. You get structure, visuals, charts, speaker notes, and editable PPT export. The site also supports PDF, Word, Excel, and web inputs. It offers templates and PDF-to-slide tools.

Search signals: Autocomplete shows dokie ai ppt and dokie ai presentation maker. The presentation lane has thin affiliate listicles. Dokie passed our gate faster than random video widgets on the same page.

Why it made the cut: June’s AIxploria “today” list put Dokie next to Aleph 2.0 and ThumbnailCreator. Slides are a repeat job: QBRs, training, pitches. Tools with real PPT export matter to office workers more than another face-swap app.

Workflow readers should try: Paste a one-page brief. Name the audience and slide count. Generate the deck. Open the export in PowerPoint. Fix fonts and brand colors. Then present. Weak prompts still make weak decks. The same rule applies to Gamma and SlideAI.

Pricing signal: AIxploria lists it as freemium. We have not published a verified credit table yet. Sign up. Run two decks. Check whether exports keep editable shapes or flatten to images.

Watch if: You need deep enterprise brand templates locked to your style guide on day one.

Skip if: You only want web-native scroll decks with no Office export. Gamma may fit better.

• Lane: Creators / slides
• Competes with: Gamma, SlideAI, Copilot in PowerPoint
• Full review: Dokie AI Review (2026) (Watch until two client exports on your templates)

---

3. Claude Opus 4.8

Verdict: Use (model default upgrade, not a new app)

Claude Opus 4.8 is Anthropic’s frontier Opus-class update on AIxploria with high upvotes. Vendor copy talks about more reliable long-form work, honesty improvements, and a fast mode with more tokens per second at the same API rate as the prior Opus revision (per directory summary). For AI Tools Radar readers, the action is not “download Opus 4.8 the app.” You update model defaults in Claude apps, API projects, and any third-party tool that exposes Anthropic model pickers.

Search signals: claude opus 4.8, anthropic opus 4.8, and best ai model 2026 rose in our June pass alongside GPT-5.5 news.

Practical jobs: Executive memos, careful editing, long analysis, contract-first drafts (still need human legal review). Pair with slide tools for decks. Opus does not replace layout.

Skip as sole stack if: Legal mandates a different vendor, or you only need the cheapest batch translation (route those jobs to smaller SKUs or DeepSeek-class APIs per policy).

How we track it: Living row on Latest AI Models Compared (2026). We bump updatedDate on that hub when release notes change, not when every radar week ships.

30-minute upgrade checklist: List your top five tasks. Note which app model name you used last month. Run one A/B prompt on Opus 4.8 if enabled. Roll back if latency or cost spikes without quality gain.

• Lane: Models (watch hub + radar callout)
• Competes with: GPT-5.5, Gemini 3.x, DeepSeek V4 (task-dependent)
• Read next: Models hub section on Claude Opus 4.8

---

4. VmakeAI

Verdict: Watch

VmakeAI helps you make UGC-style product videos for TikTok, Reels, and Shopify. You do not need a full studio setup on day one. You can start from a product link, a photo, or a short idea. The tool bundles AI avatars, automatic captions, up to 4K upscaling, background removal, and visual cleanup in one pipeline. That is attractive for small shops that want many ad variants fast. AIxploria lists it as freemium under Video Generators and Marketing.

Why Watch, not Use yet: The feature list looks strong on paper. We still need proof before we call it worth trying for everyone. Open questions include export rights, watermark rules, avatar likeness terms, and music licensing. Directory pages rarely say whether you can run paid Shopify ads with the default audio track. Synthetic spokespeople can trigger platform policy reviews. We want one full test run and a clear read of Terms before we upgrade the verdict.

Test plan (publish when done): Build one product clip from a public URL. Record one avatar script read. Note caption accuracy and timing. Check if 4K export needs a paid plan. Save screenshots of the billing page. Read Terms for commercial ad use.

Best for (pilot audience): Solo merchants running cheap creative tests. Performance marketers who have in-house legal review before spend goes live.

Skip if: You need broadcast-clear talent releases on day one. Skip if you will not read billing before you promise a client a clean export.

Competes with: InVideo, Kling-style generators, Grok Imagine, and UGC ad suites in the same directories.

• Lane: Creators (video)
• Full compare: Kling AI 3.0 vs Grok vs Veo (2026)

---

5. WebZum

Verdict: Watch

WebZum showed up on TAAFT trending in our June scan (listed as WebZum 2.7 on directories; we verify the product at webzum.com). It is a chat-based website builder. You describe what you want. You can upload photos. The tool returns a site draft in minutes. That fits our builders lane for readers who do not want to open an IDE. The trade-off is messy search demand. Google suggest still mixes “WebZum” with “websim” and similar names. We track buzz, but attribution is harder than for a unique brand like Dokie.

Why it made the radar: June’s AIxploria “today” feed was packed with video and slide tools. Builders were thin. WebZum is our experiment slot for the question: “Can AI ship a landing page this week?” Early demos often look polished. Real projects need forms, analytics, and mobile checks.

What to verify yourself: Custom domain support and DNS steps. Mobile layout on a real phone, not only the preview pane. Contact forms and analytics hooks. Whether images and fonts are licensed for commercial use. Export and hosting lock-in if you leave the platform. Run Lighthouse on the published URL. Chat builders often score pretty in demo and weak on performance.

Best for: Founders who need a first draft to test an offer. Event pages and simple portfolios are fair game. Treat output as a starting point, not the final brand site.

Skip if: You need regulated ecommerce, complex logins, or a CMS your team already runs on Webflow or WordPress.

• Lane: Builders
• Full review: Lovable AI Review (2026)
• Note: Re-test when webzum.com posts public pricing changelog

---

6. Aleph 2.0

Verdict: Watch

Aleph 2.0 is on AIxploria as frame-level video editing. Change color, clothing, or background on one frame and spread the edit across the clip (up to 30 seconds, 1080p). Untouched regions stay stable. Directory links point to Runway studio access (app.runwayml.com edit mode). Treat Aleph as part of the pro video stack, not a casual mobile filter app.

Why Watch: Niche but real for short-form editors fixing wardrobe continuity, background cleanup, or product color swaps without reshooting. Paid tier on directories; we have not finished cost-per-clip math for typical 15s ads.

Best for: Performance editors already on Runway or willing to pay for precision edits on short clips.

Skip if: You need full-length episode retiming or legal clearance on likeness edits for real actors without contracts.

Competes with: CapCut AI tools, Adobe Premiere AI features, HeyGen for avatar-led clips (different job).

Editor note: AIxploria labels Aleph as paid and verified. Budget one or two test clips before you commit a client campaign. If flicker appears on edges, fall back to traditional rotoscope for hero shots.

• Lane: Creators (video edit)
• Compare: Kling AI 3.0 vs Grok vs Veo (2026)
• Test note: Run one 10s clip with a single background swap, inspect temporal flicker on hair and hands

---

7. SlideAI (slideai.net)

Verdict: Use (presentation drafts)

SlideAI is an AI presentation generator (v2.0). You give a topic or paste notes. It returns structured slide copy: headings, bullets, and a logical flow. You edit before you worry about design polish. Optional web research can pull recent public stats when you toggle it on. Always fact-check numbers before executive meetings. Free daily credits apply on the free tier. Paid packs start at $1.99 for 20 credits as of June 2026 per our review.

Why it is worth trying for drafts: The blank slide problem is real. Consultants, teachers, and founders often stare at slide one for too long. SlideAI gets you to a rough deck fast if your prompt names audience, goal, and scope. The site is honest that human review is expected. That matches real-world Gamma tests too. AI gets you partway. You still own trust, tone, and brand.

Good prompt habit: Write who is in the room, what decision you want, and how many slides you need. Skip research for internal training decks. Turn it on for market updates where fresh stats help. Export, then fix charts, fonts, and logos in PowerPoint or Google Slides.

Not for: One-click keynote design with zero edits. Pixel-perfect enterprise templates without your design system. Anyone who will not sign in to buy credits when the free daily cap runs out.

Workflow: Specific topic, pick slide count, generate, export, polish in Office. Compare exports with Dokie this week if PowerPoint-native output is your main filter.

Competes with: Dokie (PowerPoint-first, featured this week), Gamma (visual web decks), Copilot inside existing Office files.

• Lane: Creators / slides
• Full review: SlideAI Review (2026)
• Cross-read: Dokie section above for export-first alternative

---

Skipped this week (logged, not forgotten)

We log skips so Week 2 does not re-discover the same tools without a reason code.

---

Next week watch list

• DeepSeek V4 comparison: DeepSeek V4 vs ChatGPT vs Claude (2026)
• OpenRouter free models: OpenRouter Free Models (2026)
• Roger (Google Ads agent) if roger google ads ai queries rise
• VidFlux if video cluster search heats up
• Dokie full review if export tests pass on two corporate templates
• Manus pricing update if vendor changes per-task billing

---

Lane balance this week

That mix matches our 2026 strategy: one agent, one builder experiment, and several creator tools because June directories were video- and slide-heavy. If you only read one section, match your job row in the lane table, then open the linked full review when available.

How this post stays fresh

We update this URL’s updatedDate and the Changelog section when facts change (pricing, access, or a verdict move from Watch to Use). When a new calendar week starts, we publish a new slug such as new-ai-tools-2026-june-week-2 instead of stuffing unrelated tools into this page. That keeps Google from mixing unrelated weekly launches on one stale URL.

Subscribe via RSS or revisit the radar index every Tuesday. For model-only updates, bookmark the Latest AI Models Compared (2026) hub, which we refresh within two weeks of major vendor releases.

---

Bottom line: Watch Manus and Dokie until your pilot matches our full reviews; Use SlideAI for fast deck copy (see review for limits); Use Claude Opus 4.8 as a model upgrade path; stay on Watch for VmakeAI, WebZum, and Aleph 2.0 until rights, hosting, and billing are verified on your account.