Find AI agents you can trust
Discover, compare, and review AI agents, vetted by real users and independently benchmarked by AgentVet Lab.
Why you can trust AgentVet
- Verification
Independent Benchmarking
Standardized testing with publicly verifiable results.
- Methodology
Five-Metric AgentVet Score
Agents are evaluated for accuracy, speed, ease of use, reliability and value.
- Integrity
Prompt-Injection Testing
Agents are tested against adversarial instructions and unsafe data-exfiltration attempts.
- Live Monitor
30-Day Provider Monitoring
Daily uptime and incident tracking for major AI providers.
- Independence
Independent Rankings
No pay-for-placement or sponsored ranking positions.
🆕 Newest Agents
Data as of November 5, 2026, refreshed live in your browser.
Larkgrass
Larkgrass gives you a team of AI agents, each with its own computer and your apps, like Gmail and Calendar. They check with you before anything is sent. Ask in plain words and it does the task: planning, booking, research, follow-ups. Repeating tasks run on a schedule. Your first job is free with no card; plans start at $25 a month.
Be the first to rate!
Free first job, no card; from $25/month
Vectoree
Give your coding agent a backend: models, auth, Postgres and storage behind one key. Skill + CLI + OpenAI-compatible gateway. Pay-as-you-go.
Be the first to rate!
Genies
Genies are AI agents that read the ERP, CRM, database and support systems you connect and act across them in one run, such as reading a Salesforce deal, checking a NetSuite order and opening a Zendesk ticket. Build in plain English, no-code or pro-code. Two-way sync gives agents current data. Approval steps hold sensitive writes until a person signs off, and every run is logged and replayable. Handles invoice intake, reconciliation, onboarding and X12 EDI. 14-day free trial.
Be the first to rate!
14-day free trial / $200/mo billed annually ($250/mo monthly) AI Agents plan
Noma travel planner
Noma travel planner lets Claude, ChatGPT or Cursor list a traveller's trips, read itineraries and saved places, and add places or itinerary items. Connect through https://grateful-horse-681.convex.site/mcp using streamable HTTP and OAuth 2.1. Users sign in to their Noma account and approve access at app.nomatravel.xyz. Booking, payments and account deletion are not exposed. 7-day free trial, then $4.99/month or $29.99/year.
Be the first to rate!
7-day free trial, then $4.99/month or $29.99/year.
aiFetchly
aiFetchly is an open-source, local-first desktop AI agent workspace for research, document processing, customer communication, scheduled workflows, plugins, and business knowledge.
Be the first to rate!
Free plan: Community FREE; Go $5/mo; Plus $10/mo; Pro $30/mo
Claude Duet
Anthropic's proactive personal agent that keeps working in the background, flags what needs your attention and checks with you before acting.
Be the first to rate!
Included with Claude Pro, Max and Ultra during research preview (US waitlist); uses plan usage limits
Featured Agents
Scite
AI research tool that analyzes scientific papers and identifies supporting or contradicting citations.
Free / $12/mo
How does AgentVet evaluate AI agents?
AgentVet.ai is the home of AgentVet Lab, an independent AI agent benchmarking and certification service. AgentVet combines structured user reviews with independent testing to evaluate AI agents for accuracy, speed, ease of use, reliability and value. Benchmark and certification results can be verified publicly.
- Step 1
Discover and compare agents
Browse the directory by category, capability and deployment type, then compare agents side by side.
- Step 2
Review standardized performance evidence
Every agent is scored on the same five metrics, so results are comparable across the directory.
- Step 3
Verify AgentVet Lab certifications
Certificates issued by AgentVet Lab carry an ID that anyone can verify publicly.
Why trust AgentVet rankings?
No pay-for-placement
Ranking positions are not for sale. There are no sponsored slots in the directory or on the rankings page.
Independent testing
AgentVet Lab runs standardized challenges against each agent instead of relying on vendor-supplied numbers.
Publicly verifiable certificates
Each certification has a public certificate ID and page, so buyers can confirm a claim without contacting us.
Multi-dimensional user reviews
Reviewers rate accuracy, speed, ease of use, reliability and value separately rather than leaving a single star count.
Clearly documented methodology
The five-metric AgentVet Score, certification tiers and prompt-injection testing are documented and applied the same way to every agent.
Popular AI Benchmark & Category Pages
🏆 Top Rated This Week
Data as of November 5, 2026, refreshed live in your browser.
Doximore
Upload PDFs, Word docs, and Excel files. Get instant, accurate answers without endless scrolling.
Free/ $9.99 per month
CardStatementHelper
Analyze your credit card statement, tips to improve your credit score, and helps you write dispute letters and submit to the dispute department of the credit card issuer.
$9.99/mo and $19.99/mo
Claude Code
Agentic coding tool by Anthropic that lives in your terminal — reads your codebase, writes and edits code, runs commands, and executes multi-step engineering tasks autonomously.
Included with Claude Pro/Max or API usage
Claude by Anthropic
Claude is an AI assistant by Anthropic that helps with writing, analysis, coding, research, and complex reasoning tasks. Known for being safe, honest, and highly capable across a wide range of professional use cases.
Free / $20mo Pro / $25mo Team
Deepgram
The most accurate and cost-effective real-time APIs for speech-to-text, text-to-speech, and voice agents — available in real-time and batch, cloud and self-hosted.
Free tier available / usage-based paid plans
Grammarly AI
AI writing assistant that improves grammar, tone, and clarity for emails, documents, and messaging.
Free / $12/mo
Join the Community
Get weekly picks, new agent drops, and exclusive reviews straight to your inbox.