SMB outbound sales teams wanting data, sequencing and AI in one tool
5M+ users doubling year over year and 4x growth in MCP teams Mar to Aug 2026.[1]
Data · AI Tools Radar · researched November 2, 2026
We track 71 AI tools businesses use today, and 53 of them are rising. The biggest shift this quarter is AI coworkers: bots like Viktor, Grok Bot and Claude Tag that take on multi-step work inside Slack, Teams and your apps. Each tool below lists what people use it for, a small-business fit rating and the watch-outs, cited to 249 vendor pages, press reports and 2026 surveys. Researched November 2, 2026.
Ask the Radar
Describe the job in a sentence or two. Claude picks one tool or a combination from the 71 on this page, with first steps and each tool's published price.
Answers use only the tools, prices and watch-outs on this page, researched November 2, 2026. Check prices on the vendor's site before you buy. Don't include personal, patient or client details.
Where we would start
Rated 5 of 5 for a company of 5 to 200 people and gaining momentum this quarter.
SMB outbound sales teams wanting data, sequencing and AI in one tool
5M+ users doubling year over year and 4x growth in MCP teams Mar to Aug 2026.[1]
Non-designers producing everyday marketing collateral
Craft launched Oct 22, 2026 for all users.[2]
Fast short-form social video
First CapCut World conference in October 2026 and a ChatGPT plugin announced.[3]
Coding, careful multi-step knowledge work and team agents in Slack
Three 5.5 models shipped in about two weeks, and Anthropic posted an S-1 IPO filing on Oct 26, 2026.[4]
Builders who want an agent doing multi-step work in a real codebase
23 versions shipped in October 2026 alone.[5]
Slack-centric teams on Claude Team or Enterprise wanting an admin-governed AI teammate per channel
Replaced the old Slack app in August and a Microsoft Teams waitlist signals expansion.[6]
Voiceover, localization and building voice AI agents
Shipped live call translation across 90+ languages with Deutsche Telekom as an early customer, weeks after Eleven v4 and Eleven Realtime.[7]
Fast pitch decks and proposals without a designer
Gamma 5 relaunch in October 2026 with 100M+ users.[8]
Marketing and sales teams making personalized or multilingual spokesperson video
HeyGen now ships its own voice and general video models, not just avatars, and is pricing them aggressively against ElevenLabs and Veo.[9]
Non-technical founders and small teams who need a real app fast
Raised $400M at a $13.3B valuation on Aug 12, 2026, nearing a $600M revenue run rate.[10]
Anyone building apps who needs a real database and auth without servers
Raised $500M at $10.5B in June 2026; platform customers up 370% in six months.[11]
SMB and mid-market teams that run their work in Slack or Teams and want one shared AI teammate
Sacra estimates about $50M annualized revenue in Sep 2026, up from a $15M run rate in May.[12]
Grouped by category, best small-business fit first. Fit is Socket.Solutions' 1 to 5 rating (draft where not yet reviewed); everything else is from the numbered sources.
The general AI assistants most teams start with.
| Tool | What people use it for | Momentum | Fit | Watch out | Business price | Sources |
|---|---|---|---|---|---|---|
| Claude Anthropic | Agentic coding, long document and analysis work, and producing docs, slides and spreadsheets. Claude Tag puts a shared Claude agent in Slack channels that teams hand work to. Latest: Claude Opus 5.5 (Sep 22), Sonnet 5.5 (Sep 28), Haiku 5.5 (Oct 7, 2026); Fable 5.1 (Sep 15); Claude Duet preview for US Pro and Max (Oct 21); Chat and Cowork unified for Enterprise with an Efficient mode (Oct 23); S-1 filed for Nasdaq IPO, ticker ANT (Oct 26, 2026) | Rising Three 5.5 models shipped in about two weeks, and Anthropic posted an S-1 IPO filing on Oct 26, 2026. | ■■■■■ 5 | Claude Duet is invite only, US only and not yet on Team or Enterprise plans, and it draws on the same usage limits as chat. The pending IPO may bring pricing and packaging changes; the S-1 also flags safety related delays as a business risk. | Team Standard $20/seat/mo annual ($25 monthly); Premium $100/seat/mo annual ($125 monthly) | [4][13][14] |
| ChatGPT OpenAI | Drafting, research and analysis for most teams. ChatGPT Work pulls context from Drive, Slack, Teams, Gmail and Salesforce to build documents, sheets, decks and web apps; Dots are always-on agents. Latest: GPT-6 Astra (Sep 2026); GPT-6.1 Sol (Sep 29) and GPT-6.1 Terra (Oct 26, 2026); GPT-Rosalind-6 biology model for vetted users (Oct 30) | Rising Three flagship models shipped Sep to Oct 2026, and OpenAI says ChatGPT serves over 1 billion users. | ■■■■■ 4 | Tiers and model names change fast (Astra, Sol, Terra, Luna, Rosalind, Pro 500); re-test workflows after each change. A share link flaw that could hijack sessions was fixed Oct 26, 2026, so treat shared chat links with care. | Not published | [15][16][17] |
| Gemini | Built into Gmail, Docs, Sheets and Meet for drafting and summaries. New small-business tools add a business hub, on-brand images, and links to Google Ads, Business Profile, Shopify, Xero and Canva. Latest: Gemini 4 Carbon (Oct 28, 2026), limited to selected cyber defenders for now; Gemini 4 Argon (Oct 13, now generally available in the app for AI Pro and Ultra); October Gemini Drop (Oct 30) adds a Windows app, Skills, file editing in chat and more connectors | Rising Two Gemini 4 models in about two weeks: Argon on Oct 13 and Carbon on Oct 28, plus small-business tools on Oct 20. | ■■■■■ 4 | Gemini 4 Carbon is not yet available to businesses and has no release date. Argon in the app is limited to AI Pro and Ultra, and Gemini 4 API prices double after the introductory period. | Workspace Business Standard with Gemini: $14/user/mo from Feb 10, 2027 ($7 promo until then) | [18][19][20] |
| Grok xAI (SpaceXAI) | Coding through Cursor and Grok Build, real-time information inside X, and image and video with Grok Imagine. Grok Bot gives always-on agent teammates that run routines on their own cloud computer. Latest: Grok 4.7 (Sep 21, 2026); Grok Bot agent teammates (beta Aug 11, 2026); SpaceX Super Intelligence Partner Network for enterprise Grok deployments (Oct 22, 2026); no Grok 5 yet | Rising Two flagship releases in six weeks (4.6 in Aug, 4.7 in Sep) and new distribution via Bedrock, Foundry and GitHub Copilot. | ■■■■■ 3 | Deepfake controversies and xAI's admission of partial distillation from OpenAI models; looser, less filtered style. | Grok Business $30/user/mo | [21][22][23] |
| Meta AI / Muse Meta | Mostly consumer use: an agent in the Muse app and WhatsApp that books travel, fills forms, shops and sends email. Not yet aimed at business teams. Latest: Muse personal agent app (Sep 8, 2026) on Muse Spark 1.3, now in the US, Canada and Singapore (Oct 28); Muse Wonder and Muse Wonder Full frontier models via Meta Model API and Muse Code (Oct 12, 2026) | Rising First Meta frontier model (Muse Wonder) shipped Oct 12, and NVIDIA agreed on Oct 27 to invest $3B in Meta's equity offering while deepening Muse inference work. | ■■■■■ 3 | Consumer focused with no business tier, available only in the US, Canada and Singapore; trains on interaction data by default. | Not published | [24][25][26] |
| Microsoft 365 Copilot Microsoft | Drafting, summarizing and email triage inside Word, Excel, Outlook and Teams. Cowork runs multi-step tasks, Code builds trackers and apps, and Autopilot is a proactive agent inside the tenant. Latest: 25+ finance skills and plugins for Excel and Cowork (Oct 29, 2026); rebuilt Copilot with Home, Code and Autopilot (Sep 25); GPT-6.1 Sol and Claude Sonnet 5.5 in the model picker | Rising Microsoft's biggest Copilot overhaul was announced Sep 25, 2026, adding multi-vendor frontier models. | ■■■■■ 2 | Billing is split: Chat is in the license, but Cowork, Code, Autopilot and frontier models use usage credits, so set spend policies. From Dec 1, 2026 Copilot can be bought without a Microsoft 365 base plan, and pay as you go is on by default for new CSP Copilot Business licenses. | Copilot Business add-on $18/user/mo annual through Dec 31, 2026 ($21 list); $25.20 monthly | [27][28][29] |
Bots that take on multi-step work in Slack, Teams, email or a browser.
| Tool | What people use it for | Momentum | Fit | Watch out | Business price | Sources |
|---|---|---|---|---|---|---|
| Claude Tag Anthropic | Anyone in a Slack channel tags @Claude to hand off multi-step work: pulling metrics, working support tickets, investigating bugs or coding. It keeps channel memory and schedules its own follow-ups. Latest: Beta for Team and Enterprise; replaced Claude in Slack on Aug 3, 2026; Teams version on waitlist | Rising Replaced the old Slack app in August and a Microsoft Teams waitlist signals expansion. | ■■■■■ 5 | Scope credentials per channel; memory persists unless Owners edit or delete it. | Consumption-based on Claude Team and Enterprise, with org and channel spend caps | [6][30][31] |
| Viktor Viktor | Teams message Viktor in Slack or Teams to pull reports, build dashboards and internal apps, run ad and marketing campaigns, and run recurring work across HubSpot, Meta Ads, Notion and QuickBooks. Latest: Side-pane link previews for GitHub PRs, Google files, Gmail threads and Slack messages (Oct 31, 2026); emoji reactions and inline video (Oct 30); email hover cards (Oct 27); per-channel usage view for admins (Oct 23); Viktor's Screen opened to all workspaces Oct 22, 2026 | Rising Sacra estimates about $50M annualized revenue in Sep 2026, up from a $15M run rate in May. | ■■■■■ 5 | Credit billing can get expensive; it has broad tool access, so set approval rules and per-person permissions. | Free (up to $100 in credits); Team $100/mo for 40,000 shared credits, no per-seat charge; Enterprise custom | [12][32][33] |
| ChatGPT Work OpenAI | Multi-hour projects that gather context across connected apps and return finished documents, spreadsheets, decks or web apps, such as turning customer research into a campaign brief. Latest: Launched Jul 9, 2026; GPT-6.1 Terra rolling out in ChatGPT Work and Codex for Plus, Pro, Business, Enterprise and Edu (Oct 26, 2026) | Rising Moved to GPT-6 within about 10 weeks of launch and is rolling across paid plans. | ■■■■■ 4 | Chat and Work can run different model snapshots; re-test workflows after model changes. | Not published | [34][35][36] |
| Town Town | A personal assistant (Townie) for each employee wired into email, docs and calendar: sales call prep, cross-team handoffs between finance, ops and field sales, shared team workflows, and emailing or texting customers, partners and candidates Latest: 1Password integration lets a Townie sign in to sites to finish tasks without Town seeing credentials (Oct 29, 2026); teammates can edit shared routines in threads (Oct 21); $108M Series B at a $1B valuation (Oct 19, 2026) | Rising Unicorn round four months after a $55M Series A; company says paying customers grew 4x and paying teams 9x since July | ■■■■■ 4 draft | Traction numbers are company claims only; credit-based pricing with overage charges, and agents that email outside parties need clear guardrails | Free $0; Starter $15/mo; Pro $49/user/mo (annual); Power $99/user/mo (annual); Power Plus $199/user/mo (annual); Team plans with pooled credits | [37][38][39] |
| Copilot Cowork and Studio Microsoft | Cowork runs long multi-tool tasks inside Microsoft 365. Copilot Studio builds no-code custom agents, voice agents and multi-agent workflows on 1,500+ connectors. Latest: Copilot Cowork GA (Jun 16, 2026); app building added Sep 10, 2026 | Rising Cowork reached GA worldwide and Microsoft ships monthly agent features. | ■■■■■ 3 | Usage-billed per task; set tenant and user spend limits and govern who can publish agents. | Copilot Studio $200 per 25,000-credit pack/mo; Cowork $0.01 per Copilot Credit plus M365 Copilot license | [40][41][42] |
| Genspark Genspark (MainFunc) | A multi-model super agent for slides, financial models, docs, websites and apps, now with long-term memory, agentic email and team chat. Latest: AI Workspace 6.0 (Jul 21, 2026) with SecondBrain memory, GenMail and GenTeam | Rising $100M Series B extension at $2.6B and 7,000+ business customers by July. | ■■■■■ 3 | Credit-heavy tasks burn through plans; memory retains company context, so review data policies. | Plus $24.99/mo; Pro $249.99/mo; Team $40/seat/mo; Enterprise custom | [43][44][45] |
| Hermes Agent Nous Research | A self-hosted, self-improving agent people reach through Telegram, Slack, Discord, WhatsApp, email or the command line. It keeps memory, runs scheduled reports and briefings, delegates to subagents, and turns solved problems into reusable skills. Latest: v0.22.5 (Oct 22, 2026), after v0.22.0 with signed desktop apps and one-click local models (Oct 13) | Rising Nous Research raised a reported $90M at a $1.5B valuation in October 2026, citing 22.7M Hermes Agent downloads. | ■■■■■ 3 draft | It has broad access to your machine and accounts and a long 2026 vulnerability list, including remote code execution through malicious repos and plugins. Keep it updated, run it sandboxed, connect few accounts and vet every skill or plugin. | Free and open source (MIT); optional Nous Portal plans: Plus $20/mo, Super $100/mo, Ultra $200/mo | [46][47][48] |
| Hone Hone | Builds AI agents meant to act as professional staff on business tasks that run for weeks or months. Latest: Launched with a $60M seed at a $285M valuation (Oct 8, 2026) | Rising Large seed from Benchmark and Index in Oct 2026, though the product is very early. | ■■■■■ 3 | Brand-new company with little public product detail; pilot only. | Not published | [49] |
| Manus Butterfly Effect (Singapore) | A general agent for research, reports, file processing, slides, video and building apps and websites from one prompt. Latest: Manus 2.0 with Manus Studio and Automations (Oct 30, 2026 post, desktop build 2.0.4); independent again after Meta unwind (Aug to Sep 2026); raised $500M+ in Oct 2026 | Rising Caixin reports ARR near $400M by June 2026, and a $500M+ round closed in October. | ■■■■■ 3 | Ownership and data handling have been unstable; review data residency for sensitive work. | Not published | [50][51][52] |
| Lindy Lindy | Inbox triage, meeting notes and follow-ups, scheduling, phone calls and recurring routines. Watch and Learn lets you show it a task once so it can take over. Latest: Lindy Apps: request internal tools built on your existing integrations (Oct 28, 2026); webhooks for Lindy Apps and full Outlook signatures (Oct 29); Airtable, Intercom and Google Analytics integrations (Oct 18) | Steady Still shipping, but now competes head-on with Claude Tag and Viktor in Slack. | ■■■■■ 3 | Team now lists at $49.99 per active user for 5,000 credits (up from $29.99 for 3,000), and monthly credits do not roll over, so cost is hard to predict. Teammate reportedly runs on DeepSeek, which regulated buyers should vet. | Free ($50 in credits for 7 days); Team $49.99/user/mo with 5,000 credits; top-ups $10 per 1,000 credits; Enterprise custom | [53][54][55] |
| Perplexity Comet Perplexity | An agentic browser that summarizes pages, queries Gmail and calendar, and completes multi-step web tasks like shopping or form filling. Latest: Comet free worldwide with Comet Enterprise (Mar 2026); Computer Automations (Oct 5, 2026); Computer adds Ultrafast models (Max) and role-based connector controls for Enterprise (Oct 19, 2026) | Steady Keeps shipping but faces Amazon litigation, with a dismissal hearing set for Nov 20, 2026. | ■■■■■ 3 | Agentic browsing can be hijacked by prompt injection; use enterprise policies and approvals. | Free to install; some features need Pro ($20/mo) or Max | [56][57][58] |
| Grok Bot xAI | Teams message Bots like coworkers to do browser and terminal work on a persistent cloud computer: prospecting and outreach drafts, inbox triage, routine refunds, meeting notes and website builds. You can teach a task once by recording it and run it on a schedule. Latest: Beta Aug 11, 2026; included with SuperGrok and Cursor plans from Aug 26; Grok 4.7 trained on the Grok Bot harness (Sep 21) | Rising xAI widened access from top plans to SuperGrok and all Cursor Teams plans in August, then trained Grok 4.7 specifically for the Bot harness in September. | ■■■■■ 2 draft | All of a user's Bots share one cloud computer, logins and files, so they are not a security boundary. No published SOC 2 or HIPAA claims and no spend cap on overages; keep regulated data out. | No standalone price; included with SuperGrok tiers and paid Cursor plans, with Bot usage metered separately and on-demand usage billed by tokens | [23][59][22] |
| OpenClaw OpenClaw Foundation (open source) | A self-hosted personal agent people message through Telegram, Signal, WhatsApp or Discord to handle email, calendars, browser tasks and scheduled jobs. Freelancers use it for lead generation. Latest: v2026.10.8 stable plus v2026.9.35 extended-stable gateway-only line with backported security fixes (both Oct 27, 2026), after OpenClaw 2.0 (Aug 30, 2026) | Rising A foundation launched July 2026 with OpenAI, NVIDIA, Microsoft and Tencent as sponsors. | ■■■■■ 2 | Serious security record: malicious skills, prompt injection and 2026 CVEs; sandboxing is off by default even in 2.0. | Free and open source (MIT); you pay for model usage | [60][61][62] |
| Salesforce Agentforce Salesforce | Prebuilt and custom agents for sales development, service, back office and supply chain inside Salesforce. Latest: Dreamforce 2026 (Sep 15 to 17): Agentforce Coworker GA; Agent Context Engine in preview (Oct 20, 2026); long-horizon agents due Nov 2026 | Steady Dreamforce shipped many features, but no published customer ROI or cost-to-run figures. | ■■■■■ 2 | Salesforce plans to charge Flex Credits for agentic MCP and API calls (counted as headless platform interactions); partners pushed back in Oct 2026 and pricing details are still unresolved, so model usage costs before rollout | Not published | [63][64][65] |
Tools that write, ship and run software and apps.
| Tool | What people use it for | Momentum | Fit | Watch out | Business price | Sources |
|---|---|---|---|---|---|---|
| Claude Code Anthropic | An agentic coding assistant in the terminal, IDE and desktop that builds features, fixes bugs, reviews code and automates repo tasks through hooks and plugins. Latest: v2.1.316 (Oct 31, 2026): pooled group budgets and a subagent effort setting; v2.1.314 (Oct 30) limits writes to working directories; Claude Mods and blocking hooks added in October | Rising 23 versions shipped in October 2026 alone. | ■■■■■ 5 | Lower-tier usage limits can bind for heavy agent work; fast releases change behavior often. Opus 4.8 fast mode retires before Nov 30, 2026. | In all paid Claude plans: Pro $20/mo; Max from $100/mo; Team Standard $25/seat monthly | [5][66] |
| Lovable Lovable | Non-developers describe an app in chat and get a working, hosted web app with database, login and integrations, used for MVPs, internal tools and customer sites. Latest: Credit run-out email warnings and one-click 2FA (Oct 28, 2026); workspace Events and Hooks and Scrive e-signatures (Oct 27); 17 releases in Oct 2026, incl. Supabase projects from chat and n8n connections (Oct 20) | Rising Raised $400M at a $13.3B valuation on Aug 12, 2026, nearing a $600M revenue run rate. | ■■■■■ 5 | Credit billing climbs on heavy iteration; generated apps need a security review before handling sensitive data. | Pro from $25/mo; Business from $50/mo (100 credits, scaling up) | [10][67][68] |
| Supabase Supabase | The Postgres backend (database, auth, storage, functions) behind many AI-built apps; Lovable and others create Supabase projects automatically. Latest: Unified .NET SDK v8 (Oct 21, 2026); scoped personal access tokens by default; Supabase Select 2026 (Oct 2): Turso acquisition and an MCP server for apps | Rising Raised $500M at $10.5B in June 2026; platform customers up 370% in six months. | ■■■■■ 5 | AI-built projects often skip row-level security; costs grow with usage. | Free; Pro from $25/mo; Team from $599/mo | [11][69][70] |
| Cursor Anysphere (SpaceX) | An AI-first code editor for writing, refactoring and reviewing whole codebases, and running agents that open pull requests and watch deployments. Latest: Agent SDK (Oct 20, 2026); Cloud Agent Canvases (Oct 16, 2026) | Rising Reportedly past $3B in annual sales, and SpaceX closed a $60B acquisition on Aug 14, 2026. | ■■■■■ 4 | New ownership, and its Composer 2 model drew criticism over undisclosed origins; pricing tiers shift often. | Free; Plus $8/mo; Premium $30/mo; Super $100/mo; Ultra $200/mo; Teams $30+/user/mo; Enterprise custom | [71][72][73] |
| OpenAI Codex OpenAI | Developers hand coding tasks from the CLI, IDE, ChatGPT or phone, run many in parallel in the cloud, and use automatic PR reviews and security scans. Latest: Codex CLI 0.166.0 (Oct 29, 2026); 0.165.0 adds a built-in Vertex AI provider (Oct 27); GPT-6.1 Terra in Codex (Oct 26); DevDay 2026 (Sep 29): Codex cloud, Code Review, Codex Security Cloud | Rising 39 CLI releases in Q3 2026 and about 25.7M weekly npm downloads, up 18.6% quarter over quarter. | ■■■■■ 4 | Features are split across ChatGPT tiers, so what a small team gets depends on its plan. | Not published | [74][75][76] |
| Replit Replit | Build and deploy full-stack apps by chatting with Replit Agent in a workspace that includes hosting, database and backups; now also investigates and fixes production alerts. Latest: Cloud Agent opens tested pull requests with screenshots and video (Oct 30, 2026); autonomous on-call engineer (Oct 23); Windows desktop preview (Oct 9) | Rising Weekly Agent releases through Sep and Oct 2026 after tripling its valuation to $9B. | ■■■■■ 4 | Usage-based agent costs can be unpredictable. | Core $20/mo; Pro $100/mo; Enterprise custom | [77][78][79] |
| GitHub Copilot GitHub (Microsoft) | Code completion and chat in the editor, plus agents that open, fix and merge pull requests or run scheduled tasks; can host Claude and Codex agent sessions. Latest: GitHub Universe (Oct 28, 2026): Copilot Fusion multi-model routing GA on all plans, shared team Copilot Automations rolling out, Copilot included with GitHub Enterprise for account-team customers; GPT-6.1 Terra GA (Oct 26) | Steady Monthly releases keep adding agent features without a major platform shift. | ■■■■■ 4 | Many agent features are previews or staged rollouts. Fusion bills each model it calls at that model's rate, and shared automations bill the repo owner's AI credits. Business and Enterprise prices were not shown on the plans page. | Free; Pro $10/mo; Pro+ $39/mo; Max $100/mo | [80][81][82] |
| v0 Vercel | Turns plain-English descriptions into React or Next.js interfaces, dashboards and full-stack apps on Vercel, often for landing pages and internal tools. Latest: v0 apps connect to Slack, Google and 100+ services (Aug 2026); v0 API GA | Steady Adds integrations and API access; SaaStr reported 4M+ users in March 2026. | ■■■■■ 4 | Best on the Vercel and Next.js stack; moving elsewhere takes work. | Free; Business $100/user/mo; Enterprise custom | [83][84][85] |
| Devin Cognition | An autonomous software engineer that takes tickets, investigates incidents and triages vulnerabilities from Slack, Teams, Jira, Linear or GitHub. Latest: Corridor security team joins Cognition to build out Devin Security (Oct 28, 2026); FrontierCode 2.0 benchmark (Oct 27); Devin Security for vulnerability scanning and fix PRs (Oct 26); $2B Series E at $48B (Sep 2026) | Rising Run-rate revenue reportedly near $900M and valuation up from $26B to $48B in months. | ■■■■■ 3 | It executes unsupervised; grant repo and production access deliberately and review its PRs. | Free; Pro $20/mo; Max $200/mo; Teams $80/mo plus $40 per dev seat | [86][87][88] |
| Jev TypeSafe AI | Developers drop Jev into apps and agent workflows as a fast decision step that returns a typed choice, score or yes/no with confidence: classifying content, routing support tickets or agent tasks, scoring leads, filtering search results and guarding LLM outputs. Latest: Jev System One decision model in early access via API and hosted MCP (launched Sep 15, 2026); in Replit AI Integrations since Oct 2 | Rising Bloomberg reported on Oct 9, 2026 that TypeSafe raised about $870M at a $7.5B valuation led by Andreessen Horowitz, weeks after launch. | ■■■■■ 3 draft | Narrow by design: it picks among options you define and cannot write text, and it can still pick the wrong one. Speed, cost and adoption claims are self-reported, and access is still through an early-access waitlist. | $0.042 per million input tokens; output tokens not billed | [89][90][77] |
| Bolt StackBlitz | Prompt full-stack web apps, slides and prototypes in the browser and deploy in one click, with visual edits and a pre-publish security scan. Latest: Bolt Forge agent on open models (Sep 14, 2026); acquired Dokai (Sep 29) | Steady Ships regularly, but with less news flow than Lovable or Replit. | ■■■■■ 3 | Opting into Bolt Forge's extra usage means sessions help train open models; check the setting. | Free; Pro $25/mo; Teams $30/member/mo | [91][92] |
Connect apps and add AI steps to repeatable workflows.
| Tool | What people use it for | Momentum | Fit | Watch out | Business price | Sources |
|---|---|---|---|---|---|---|
| Zapier Zapier | Connect thousands of SaaS apps with Zaps and add AI steps or agents, plus chatbots, Tables and Zapier MCP so AI assistants can act across apps. Latest: Standalone Agents folded into AI steps inside Zaps (Sep 2026); Functions retired Sep 1 | Steady Consolidating products rather than launching new lines. | ■■■■■ 5 | Converted agents need auditing; task billing and MCP calls add up. | Free (100 tasks); Professional from $19.99/mo; Team from $69/mo | [93][94][95] |
| n8n n8n GmbH | Visual workflows and AI agents connecting apps, APIs and models, self-hostable for data control; popular for lead routing, document processing and internal agents. Latest: n8n 2.45.4 stable (Oct 26, 2026); n8n 3.0.0 and 2.46.0 pre-releases (Oct 27); Browser Use available to all (Oct 6) | Rising SAP's May 2026 investment valued it at about $5.2B, with weekly releases through October. | ■■■■■ 4 | Self-hosting brings security and update duties; frequent breaking node changes. n8n 3.0 is a major release with breaking changes, so run the migration tool before upgrading. | Starter EUR 20/mo; Pro EUR 50/mo; Business EUR 667/mo (annual) | [96][97][98] |
| Make Make (Celonis) | Visual multi-step scenarios and AI agents that move data between apps, used for invoice lookups, CRM updates and marketing ops. Latest: Process Intelligence (task and process mining) unveiled at Waves '26, waitlist (Oct 20, 2026); MakeOS vision (waitlist); agent prompt caching (Sep 2026) | Steady Repositioning around governed agents, but MakeOS is still at the waitlist stage. | ■■■■■ 4 | AI features use more credits per action; MakeOS is a vision, not a shipped product. | Not published | [99][100][101] |
| Gumloop Gumloop | Non-technical staff build AI agents across company tools for CRM updates, lead research, support triage and data analysis. Latest: v13.3.0 (Oct 30, 2026): publish agents to Okta, native context compaction; v13.2.0 (Oct 28): App Rules to block or require approval for tool calls; v13.0.0 revamped agent harness (Oct 23) | Rising 21+ versions Aug to Oct 2026 after a $50M Series B led by Benchmark. | ■■■■■ 3 | Very fast change; fewer long-standing integrations than Zapier. | Pro from $37/mo; Enterprise custom | [102][103][104] |
| Airtable AI Airtable (Bending Spoons) | A no-code database and app builder with AI fields, AI app building and agents for tracking, operations and light CRMs. Latest: Bending Spoons acquisition completed (Sep 2026); large layoffs Oct 12, 2026 | Cooling Sold for $1.285B enterprise value, far below its $11.7B peak, then cut much of the team. | ■■■■■ 3 | Roadmap and support risk after ownership change; watch for price increases. | Team $20/user/mo; Business $45/user/mo (annual) | [105][106][107] |
Video, voice, images, music and decks.
| Tool | What people use it for | Momentum | Fit | Watch out | Business price | Sources |
|---|---|---|---|---|---|---|
| Canva AI Canva | Social posts, flyers, presentations and brand assets from templates, now with on-brand image generation that turns into editable layered designs. Latest: Canva Craft, its first in-house image model (Oct 22, 2026) | Rising Craft launched Oct 22, 2026 for all users. | ■■■■■ 5 | Check commercial-use terms for AI images. | Not published | [2] |
| CapCut ByteDance | Edit short-form social video with templates, auto-captions and effects, and generate AI clips for TikTok, Reels and Shorts. Latest: ChatGPT plugin announced Oct 24, 2026 (coming soon); Seedance video models | Rising First CapCut World conference in October 2026 and a ChatGPT plugin announced. | ■■■■■ 5 | ByteDance ownership raises data and regulatory concerns; AI generation features vary by market. | Not published | [3][108] |
| ElevenLabs ElevenLabs | Lifelike voiceovers, audiobooks, dubbing and voice cloning, and increasingly phone and web voice agents that handle customer calls in 90+ languages. Latest: SFX v3 sound effects model (Oct 29, 2026); Eleven Realtime Translation alpha for enterprise API customers (Oct 21); Eleven Realtime for voice agents (Oct 20); Eleven v4 TTS (Sep 28) | Rising Shipped live call translation across 90+ languages with Deutsche Telekom as an early customer, weeks after Eleven v4 and Eleven Realtime. | ■■■■■ 5 | Voice-cloning consent and impersonation risk; disclose AI voices on customer calls. | Free; Starter $6/mo; Creator $22/mo; Pro $99/mo; Scale $299/mo; Business $990/mo | [7][109][110] |
| Gamma Gamma | Turns a prompt, notes or a document into a polished deck, one-pager or web page in minutes, pulling context from tools like Slack and Granola. Latest: API keys and REST API opened to all plans including Free (Oct 22, 2026); Gamma 5 (Oct 6, 2026): rebuilt agent, freeform editor, about 20 connectors, PowerPoint import | Rising Gamma 5 relaunch in October 2026 with 100M+ users. | ■■■■■ 5 | Output can look generic and needs fact-checking; connectors raise data governance questions. | Not published | [8][111][112] |
| HeyGen HeyGen | Talking-avatar videos without filming for sales outreach, training, explainers and social; translating and lip-syncing videos into other languages. Latest: HeyGen Voice Turbo, ranked first on Artificial Analysis TTS board, free API tier through Nov 30, 2026 (Oct 21); HeyGen Video general video model with audio (Sep 30, 2026); HeyGen Voice (Oct 9) | Rising HeyGen now ships its own voice and general video models, not just avatars, and is pricing them aggressively against ElevenLabs and Veo. | ■■■■■ 5 | Avatars and voice clones need explicit consent of the person shown. | Free; Creator $29/mo; Pro $49/mo; Business $149/mo | [9][113][114] |
| Synthesia Synthesia | Turns scripts and documents into avatar-led training, onboarding and internal comms videos, now with interactive roleplay and survey sessions. Latest: Sessions interactive avatars (Oct 1, 2026) and new plan structure | Rising $4B valuation in January 2026, a new NYC HQ in September, and 140% net revenue retention. | ■■■■■ 4 | Custom avatars of real employees need consent and governance. | Basic free; Starter $19/mo; Pro $89/mo; Enterprise custom | [115][116][117] |
| Google Flow and Veo | Generate and edit short cinematic clips, B-roll and ad concepts from text or images, extend scenes, and make clips inside Vids for marketing and internal video. Latest: Gemini Omni 1.1 Flash in Flow (Aug 2026) and in Google Vids (Sep 23, 2026) | Rising Omni 1.1 Flash reached Flow in August and Workspace Vids in September. | ■■■■■ 3 | Outputs carry SynthID watermarks; usage is credit-limited by plan. | Not published | [118][119] |
| Higgsfield Higgsfield AI | Ad creatives, product videos and short social clips from text and images, with multi-scene agent workflows and storyboards that combine several third-party video models. Latest: Nvidia and TCV joined a Series B extension (Oct 23, 2026), after a $400M Series B at a $5.4B valuation led by DST Global (Aug 2026) | Rising Reported annualized revenue near $700M in August 2026, mostly from businesses, and Nvidia and TCV invested on Oct 23; it cites 40M+ registered users. | ■■■■■ 3 draft | Revenue figures are company-reported run rates; it depends on third-party models it does not own; past controversies include stock footage passed off as AI output and throttled unlimited plans. Check likeness and logo use before publishing. | Not published | [120][121][122] |
| Kling AI Kuaishou | Realistic product, ad and story clips up to 30 seconds guided by reference images and keyframes. Latest: Kling unit completed a funding round of about $2.98B (Oct 30, 2026); Kling 4.0 fully available on web and app (Oct 15 to 16), native 30 second video; API still early access | Rising Kling 4.0 reached general release, Kling reports 100M+ users and 50,000+ enterprise customers, and it closed a near $3B funding round on Oct 30. | ■■■■■ 3 | China-based vendor; review data residency and likeness rules. Kuaishou keeps about two-thirds of the equity, so the parent's priorities and regulators apply. | Not published | [123][124][125] |
| Suno Suno | Full songs, jingles and background music from a text prompt for videos, ads and social posts. Latest: Suno MCP beta (Oct 28, 2026) lets ChatGPT, Claude and other MCP clients create songs on any Suno plan using account credits; Suno v6 model family (Sep 9, 2026) trained on licensed partner data; answered UMG and Sony second lawsuit (Oct 20, 2026) | Rising Valued at $5.4B after raising over $400M in June 2026, with 100M+ users. | ■■■■■ 3 | UMG and Sony are suing again over v6, and in its Oct 20 answer Suno admitted it has not deleted all recordings used in past training. Discovery is escalating, with a Dec 1 hearing on an Apple subpoena. Sony and UMG have no license deal, an artist class action is pending, and Free plan songs are non-commercial. | Free (no commercial rights); Pro $8/mo; Premier $24/mo | [126][127][128] |
| Descript Descript | Record, then edit video and audio by editing the transcript; Underlord removes filler words, cuts clips and adds captions. Latest: Oct 12, 2026 roundup: ElevenLabs v4 voices, drive search, recording requests for Enterprise, file repair, faster default AI video model; Sep 17 update added AI music and sound effects, rebuilt captions and Underlord agent | Steady Regular feature updates but no major funding or model news. | ■■■■■ 3 | Voice cloning needs speaker consent. | Free; Hobbyist $24/mo; Creator $35/mo; Business $65/mo (monthly billing) | [129][130][131] |
| Midjourney Midjourney | High-style concept art, mood boards, ad visuals and brand imagery, now with instruction-based editing and short animation. Latest: V8.2 (Jul 24, 2026); V8.2 edit model (Aug 27); Thinking Mode alpha (Oct 8) | Steady Frequent editor and alpha updates, with ongoing studio copyright suits. | ■■■■■ 3 | Disney, Universal and Warner Bros. Discovery lawsuits; avoid recognizable characters or IP. | Not published | [132][133] |
| Grok Imagine xAI | Quick images and short video clips with audio inside Grok and X for social content and memes; xAI is pushing it toward long-form stories. Latest: Grok Imagine Video 1.5 used for an AI 'Odyssey' series (Oct 23, 2026); older grok-imagine-image-quality API model retired Nov 2, 2026 in favor of grok-imagine-image-2.0 | Steady The Odyssey release drew attention, but reviewers criticized its quality. | ■■■■■ 2 | Serious legal and reputational risk: deepfake history and Illinois biometric privacy class actions; never upload real people's photos. | Not published | [134][135][136] |
Cited research and answers grounded in your own documents.
| Tool | What people use it for | Momentum | Fit | Watch out | Business price | Sources |
|---|---|---|---|---|---|---|
| Gemini Notebook | Upload documents, PDFs and recordings and get answers grounded in those sources, plus audio and video overviews, slides and quizzes for briefings and training. Latest: Renamed from NotebookLM (Jul 16, 2026); Connected App in Gemini (Oct 23); mobile recorder (Oct 27) | Rising Steady Aug to Oct 2026 releases; Google cites 30M+ users and 600,000 organizations. | ■■■■■ 4 | The rename may confuse staff; check Workspace data settings before uploading sensitive or PHI content. | Not published | [137][138][139] |
| Perplexity Perplexity AI | Fast cited research for market, vendor and competitor questions, deep research reports, and a Computer agent for multi-step work that can now run on local files or in regulated environments. Latest: Perplexity Enterprise earned FedRAMP Moderate (Oct 28, 2026); Computer agent runs fully on-device on Mac and on air-gapped servers (Oct 20) | Rising Reported annualized revenue topped $750M, up from under $250M at the start of 2026. | ■■■■■ 4 | Web traffic has slipped since Oct 2025; important claims still need human review. | Pro $17/mo and Max $167/mo billed annually; Enterprise per seat, not listed | [140][141][142] |
Notetakers, CRM agents and prospecting.
| Tool | What people use it for | Momentum | Fit | Watch out | Business price | Sources |
|---|---|---|---|---|---|---|
| Apollo AI Apollo.io | A B2B contact database with AI that recommends who to contact and when, drafts and sends sequences, and works with Claude and ChatGPT through MCP. Latest: AI GTM System at ApolloNEXT (Sep 30, 2026); data partner for Salesforce's Hunter agent | Rising 5M+ users doubling year over year and 4x growth in MCP teams Mar to Aug 2026. | ■■■■■ 5 | Some announced pieces are not yet available; manage credits and deliverability. | Not published | [1][143] |
| GoHighLevel HighLevel | Agencies and local service businesses run it as a white-label CRM, funnel and messaging hub, then add the AI Employee so Voice AI answers and qualifies calls, Conversation AI replies across SMS, social and chat, and agents book appointments, chase reviews and run follow-ups. Latest: LevelUp 2026 in Dallas (Oct 26 to 29): $497 Agency Pro renamed the Founder Plan (same price) and a $20/mo per sub-account AI Employee Starter launched for Founder Plan agencies; AI Studio and Funnel AI generally available (Sep 2026) | Rising LevelUp (Oct 26 to 29) shipped a renamed Founder Plan and a cheaper $20 AI Employee Starter tier, after AI Studio went generally available in September. | ■■■■■ 4 draft | The base price understates cost: AI Employee is a per-sub-account add-on with usage fees for AI, voice and messaging, and the $20 Starter tier is only for Founder Plan agencies. Setup is heavy without an agency partner, and HIPAA costs an extra $297/mo unless you are on Enterprise. | Starter $97/mo; Unlimited $297/mo; Founder $497/mo; AI Employee add-on $50 or $97/mo per sub-account plus usage; HIPAA add-on $297/mo (included with Enterprise) | [144][145][146] |
| Read AI Read AI | Notetaker for Zoom, Teams and Meet that also uses email, calendar and CRM context; cc Ada to schedule and draft replies. Latest: Ada email digital twin (2026); available as a Claude connector, ChatGPT app and MCP server | Steady 5M monthly active users, but no new funding found for Aug to Oct 2026. | ■■■■■ 4 | Ada acts on email threads; set consent and approval rules. | Free; Pro $19.75/user/mo; Enterprise $29.75/user/mo (monthly) | [147][148] |
| Clay Clay | Combines CRM, call and product data with outside signals like funding and hiring; AI agents research prospects and set up outreach. Latest: Claygent auto-mode model routing, reported 47% cheaper than always using the top model (Oct 27, 2026); $115M Series D at $7.1B (Sep 9, 2026); Sculpt conference (Oct 8) | Rising Valuation more than doubled to $7.1B with 17,000+ customers. | ■■■■■ 3 | Credits climb quickly; complex without a dedicated operator. | Launch $167/mo; Growth $446/mo; Enterprise custom | [149][150][151] |
| Granola Granola | Transcribes meetings in the background without a bot and turns rough notes into clean summaries; teams share notes and pipe them into Claude or ChatGPT. Latest: Free voice typing in any app for all users (Oct 29, 2026); $125M Series C at $1.5B (Mar 2026); Spaces, APIs and MCP server; Apple Watch app (Jul 2026) | Rising Valuation rose from $250M to $1.5B in under a year, and on Oct 29 it moved beyond meetings with free dictation in any app. | ■■■■■ 3 | AI meeting notes are commoditizing; check API access before building on it. | Basic $0; Business $14/user/mo; Enterprise $35/user/mo | [152][153][154] |
| HubSpot Breeze HubSpot | AI agents in the HubSpot CRM that prep reps, log meeting notes, update deals, find prospects, answer customers and build campaigns. Latest: Agent Hub live; HubSpot Work in public beta; AI spend caps (Oct 15, 2026) | Rising A wave of agents in Oct 2026 plus spend controls. | ■■■■■ 3 | Credit use is hard to forecast; set account and feature caps. | Sales Hub Professional $90/seat/mo; Prospecting Agent $1 per lead; Customer Agent $0.50 per resolution | [155][156][157] |
| Fathom Fathom (Superhuman) | Records calls, flags key moments and action items, makes meetings searchable and syncs results to CRMs. Latest: Now part of Superhuman (deal announced complete Sep 14, 2026); same app, login and plans so far; Fathom connector live in Superhuman Go | Steady Hundreds of thousands of companies use it; now part of Superhuman, which plans to combine meeting context with email and docs. | ■■■■■ 3 | The deal is done, so expect roadmap and terms to shift toward Superhuman; avoid long commitments and know how to export your recordings. | Free; Premium $20/user/mo; Team $19/user/mo; Business $34/user/mo | [158][159][160] |
| Otter Otter.ai | Transcripts with speaker labels, one-page summaries and a searchable archive, plus sales agents for follow-ups and demo booking and an MCP server. Latest: Work Mode briefs and follow-ups for all paid users plus Live Assist real-time coaching in Zoom, Teams and Meet (Oct 27, 2026); $320M revenue-based financing from General Catalyst's Customer Value Fund (Oct 20) | Rising The $320M raise funds tripling sales and marketing; 40M+ users. | ■■■■■ 1 | Expect aggressive enterprise upselling; discounted prices carry conditions. | Basic free; Pro $16.99/user/mo; Business $30/user/mo (monthly) | [161][162][163] |
Ambient documentation, coding and EHR-native AI.
| Tool | What people use it for | Momentum | Fit | Watch out | Business price | Sources |
|---|---|---|---|---|---|---|
| Abridge Abridge | Ambient AI that listens to patient visits and drafts structured notes in the EHR, now extending to coding, CDI and claim review. Latest: VA enterprise contract (Sep 22, 2026); UT System agreement (Oct 8, 2026) | Rising Deployed at 300+ US health systems and selected by the VA. | ■■■■■ 4 | Enterprise-only; Epic's native Art now competes directly. | Not published | [164][165][166] |
| Ambience Healthcare Ambience Healthcare | Ambient documentation with real-time coding and CDI prompts across clinic, inpatient and ED, now extending to nursing chart Q&A and handoffs. Latest: Ambience Chorus agent harness (Aug to Sep 2026); nursing suite (Jun 2026) | Rising KLAS/CHIME Trailblazer Award and Chorus launched Aug to Sep 2026. | ■■■■■ 4 | Enterprise-only with outcome-tied fees; competes with Epic's native tools. | Not published | [167][168] |
| OpenEvidence OpenEvidence | Clinical question answering grounded in peer-reviewed literature, now with patient context inside the Epic chart at partner health systems; point-of-care evidence lookup and trainee education. Latest: Embedded in NYU Langone Health's Epic EHR across about 12.4M outpatient visits a year (Oct 29, 2026); Johns Hopkins Medicine collaboration (Oct 13, 2026) | Rising Large health systems are wiring it into Epic: NYU Langone on Oct 29 after Johns Hopkins on Oct 13. | ■■■■■ 4 draft | Inputs to the public tool are not covered by HIPAA, so keep patient identifiers out unless you use a health system integration; usage claims are company figures. | Free for verified US health care professionals | [169][170][171] |
| Suki Suki | Ambient scribe plus voice assistant that drafts notes, suggests codes and handles voice commands, integrated with Epic, Oracle Health, athenahealth, MEDITECH and WellSky. Latest: $90M Series E (Oct 15, 2026); athenahealth preferred ambient partner | Rising Fresh $90M round in Oct 2026 to grow into a clinical AI platform. | ■■■■■ 4 | Ambient scribing is commoditizing; ask for specialty accuracy data and total cost per clinician. | Not published | [172][173] |
| Microsoft Dragon Copilot Microsoft | Dictation plus ambient note capture with a clinical assistant across EHRs; partner agents for prior authorization, revenue cycle and decision support. Latest: Apps and agents GA on Microsoft Marketplace (Aug 2026); per-encounter billing from May 1, 2026 | Steady 100,000+ clinicians use it daily and Marketplace purchasing opened in August. | ■■■■■ 3 | Pricing not published; competes with native Epic AI in Epic shops. | Not published | [174][175] |
| Epic AI (Art, Emmie, Penny) Epic Systems | AI built into the EHR: Art summarizes charts and drafts notes, Emmie handles patient messaging in MyChart, and Penny automates coding and denial appeals. Latest: Connectors announced at Open@Epic (Oct 21, 2026), letting outside apps plug into Art, Emmie, Penny and Agent Factory agents; Ergo AI platform available at UGM (Aug 2026); Agent Factory in 2027 | Rising AI in Epic assists 1.4M+ clinicians monthly; Art summaries at 300+ organizations. | ■■■■■ 2 | Developer-built Connectors are not expected in Epic's Showroom until March 2027, and observers note the MCP-based design treats outside apps as background tools behind Epic's own agents. | Not published | [176][177][178] |
Models you can host yourself for private or regulated data.
| Tool | What people use it for | Momentum | Fit | Watch out | Business price | Sources |
|---|---|---|---|---|---|---|
| DeepSeek DeepSeek | Low-cost API access and self-hosted coding and reasoning. The flagship needs heavy multi-GPU hardware to run in-house. Latest: DeepSeek V4 Pro open weights (Aug 12, 2026, MIT); V4.1-Flash-1017 update with weights on Hugging Face (Oct 17, 2026); Search and Fetch APIs opened to all developers | Rising V4 Pro shipped under MIT in August, then V4.1-Flash in September with API price cuts. | ■■■■■ 4 | Search and Fetch API calls start billing at $0.75 per 1,000 calls per endpoint on Oct 31, 2026, so check agent tool usage before then. A roughly five hour outage hit the service the week of Oct 12. | Not published | [179][180][181] |
| Mistral Mistral AI | Data-sovereign and self-hosted AI for European companies, document and OCR processing, and multimodal reasoning. Latest: Mistral Large 4 open weights posted on Hugging Face (Oct 29, 2026): about 1T parameters, 1M token context, FP8 and NVFP4 checkpoints, under a modified MIT license; Voxtral 3 audio models and Voice Agents in public preview (Oct 20, 2026) | Rising Large 4 weights shipped Oct 29 with same day support in Transformers, vLLM and llama.cpp. | ■■■■■ 4 | Internal use is free, but running Large 4 as a hosted service for third parties needs a separate commercial license. Self hosting is heavy: FP8 weights are roughly 1 TB and examples use 8 GPUs. | Not published | [182][183][184] |
| Kimi K3.1 Pro Moonshot AI | Long multi step coding agents, office and financial analysis, front end and CAD scripting, long document work Latest: Kimi K3.1 Pro and K3.1 Pro Highspeed (Oct 23, 2026), 3T class, 1M token context; full open weights promised by Nov 4, 2026 | Rising New flagship landed Oct 23 with vendor benchmarks close to Claude Opus 5.5 on terminal coding, wide press pickup and same day OpenRouter listing. | ■■■■■ 3 draft | Open weights and license were still not posted as of Nov 1, 2026 (promised by Nov 4). Moonshot plans to retire all K2 models by the end of December 2026. Benchmarks are vendor reported; reasoning cannot be turned off; enterprise admins must enable it manually; data residency review needed for a China based vendor. | Not published | [185][186][187] |
| Qwen Alibaba | Self-hosted and fine-tuned deployments. The 27B model fits on a single server; the large model powers coding agents. Latest: Qwen3.8 family (Aug 2026), including Qwen3.8-27B under Apache 2.0 (Aug 14) | Steady Qwen3.8 shipped in August 2026, with no newer major release confirmed for Sep to Oct. | ■■■■■ 3 | The large open model needs a commercial license above $50M in revenue; hosted Max versions are proprietary. | Not published | [188] |
| gpt-oss OpenAI | On-premises or private reasoning. The 20b model runs on about 16GB of memory and the 120b on one 80GB GPU. Latest: gpt-oss-120b and gpt-oss-20b (Aug 2025, Apache 2.0); no successor confirmed | Steady Still the original Aug 2025 pair, with no successor confirmed. | ■■■■■ 3 | Now well behind 2026 open models like DeepSeek V4 and Qwen3.8. | Not published | [189][190] |
| Llama / Muse Glimmer Meta | Glimmer is a compact open model that runs in under 20GB for local or private use. Existing Llama installs keep running, but no new Llama models are coming. Latest: Llama line ended Apr 2026; Muse Glimmer 30B (Apache 2.0) released Aug 10, 2026 | Cooling Meta moved frontier work to closed Muse models and promised open weights had not shipped by mid-August. | ■■■■■ 3 | Plan migrations off Llama; Meta's future open releases are uncertain. | Not published | [191] |
Dated launches, funding rounds, acquisitions and shutdowns, newest first.
| Figure | Value | Source |
|---|---|---|
| US businesses on Ramp paying for any AI | 55.9% | Ramp AI Index, Sep 2026[192] |
| Ramp businesses paying for Anthropic | 43.8% | Ramp AI Index, Sep 2026[192] |
| Ramp businesses paying for OpenAI | 40.6% | Ramp AI Index, Sep 2026[192] |
| Organizations using AI in at least one business function | 88% | McKinsey: The state of AI in 2026, Jun 2026[193] |
| Large organizations scaling AI agents | 40% | McKinsey: The state of AI in 2026, Jun 2026[193] |
| Respondents reporting any EBIT impact from AI | 37% | McKinsey: The state of AI in 2026, Jun 2026[193] |
| US employees using AI at work a few times a year or more | 52% | Gallup: Artificial Intelligence indicator, May 2026[194] |
| US employees using AI a few times a week or more | 30% | Gallup: Artificial Intelligence indicator, May 2026[194] |
| US employees using AI daily | 15% | Gallup: Artificial Intelligence indicator, May 2026[194] |
| US adults who use AI chatbots | 49% | Pew Research Center, Feb 2026[195] |
| US adults who use ChatGPT | 44% | Pew Research Center, Feb 2026[196] |
| US adults who use Gemini | 24% | Pew Research Center, Feb 2026[196] |
| US adults who use Microsoft Copilot | 17% | Pew Research Center, Feb 2026[196] |
| US adults who use Claude | 6% | Pew Research Center, Feb 2026[196] |
| US firms with 250+ employees using AI | 37% | US Census Bureau: Business Trends and Outlook Survey, May 2026[197] |
| US businesses using AI to produce goods or services | 19.8% | US Census Bureau: Business Trends and Outlook Survey, May 2026[197] |
| Canadian businesses using AI | 19.2% | Statistics Canada, Q2 2026[198] |
| ChatGPT monthly web visits (No. 1 AI tool) | 5.30 billion | Business Today, Jun 2026[199] |
| Gemini monthly web visits (No. 2) | 1.14 billion | Business Today, Jun 2026[199] |
| Claude monthly web visits (No. 3) | 967.91 million | Business Today, Jun 2026[199] |
Every Monday a scheduled research run re-checks each tool's vendor pages and recent press, adds tools with real traction and rewrites this page. A fact is kept only when the run read the page that supports it; prices appear only when a vendor publishes them. Descriptions are in our own words.
Small-business fit (1 to 5, for a company of 5 to 200 people) is rated by Socket.Solutions. A rating marked draft was proposed from the cited sources and has not been reviewed yet. Momentum (rising, steady, cooling) is a Socket Solutions judgment grounded in the cited sources. This is not an endorsement. The starter list was the Nashville AI Consulting AI tools scoreboard (2026-06).
Researched November 2, 2026; next weekly refresh by November 16, 2026.
There are 71 tools on this list. Fifteen minutes to find the one that fits the job in front of you.