Agent Templates for every business function
Marketing, sales, HR, finance, and operations - pick a proven template instead of prompting from scratch. Each template encodes best practices your team can reuse.
30+ pre-built AI Agents. Ready to Deploy.
No building required. Choose a template, configure, and run.
Key capabilities
Human-in-the-Loop approval workflows
Sensitive sends and data-changing actions pause for review. Approvers see full context, edit drafts inline, and release - or reject - with a clear audit trail.
Multi-model: Claude, GPT-5, Gemini, Mistral (switch per agent)
Pick the best model per agent or per workflow step. Route high-stakes work to your preferred vendor, keep experiments on cost-efficient models, and avoid lock-in.
Real-time streaming execution
Watch agents work as they stream tokens, tool calls, and intermediate reasoning. Fewer black boxes, faster debugging, and quicker trust from stakeholders.
Interactive preview
Where agent runs meet your team
Chats and agent runs show up in one stream: pick a conversation, inspect the thread, and use @ to pull in agents or workflows — then compare with the screenshots below.
AgentWorks
AgentWorks
Chats
Board
SS-05
Agent template marketplace - browse and launch templates
1440 × 900pxSS-06
Agent detail - configuration, model, and safeguards
1440 × 900pxSS-07
Human-in-the-loop approval queue and review
1440 × 900pxFAQ
Frequently asked questions
- What is an AI agent in AgentWorks?
- An AI agent in AgentWorks is a pre-configured workflow that calls a large language model (GPT, Claude, Gemini, or Mistral) with specific instructions, tools, and access to your knowledge base. AgentWorks ships 30+ agent templates for marketing, sales, support, HR, finance, and operations, plus custom agents built by AgentWorks AI engineers.
- Can I run AI agents on ChatGPT, Claude, and Gemini in one workspace?
- Yes. AgentWorks is multi-LLM by design: every agent and chat lets you switch between Anthropic (Claude), OpenAI (GPT-5), Google (Gemini), and Mistral within the same thread. Per-model token pricing is billed transparently from a single euro wallet.
- Are AgentWorks AI agents EU AI Act compliant?
- AgentWorks is EU AI Act-ready rather than blanket "compliant" — whether a use case is high-risk depends on how you deploy the agent. Every agent can be risk-classified per the EU AI Act, runs through PII redaction at the gateway before any third-party model sees data, writes a full audit log, and supports human-in-the-loop approval for high-risk decisions. EU data residency is available on managed cloud, with private-cloud or self-host on Enterprise.
- How is pricing calculated for AI agent runs?
- AgentWorks uses transparent per-token pricing pulled directly from each model vendor; tokens are billed at cost + 10%. The wallet shows live euro spend per chat turn or agent run. Tier prices today: Starter free with €5 included, Pro €49/month, Team €49/seat/month, Enterprise custom.
- How do I deploy a custom AI agent?
- AgentWorks supports two paths: no-code custom agents you build yourself (Pro and up), and engineer-built advanced agents we develop to spec (Enterprise / done-for-you). The same wallet, audit log, and EU AI Act controls apply. Start free at app.agent-works.ai, or contact us for an advanced build.
Explore More Features
See how other platform capabilities work together.
Last updated: March 2026