- Dust is open source — is that not a deal-breaker for AgentWorks?
- For some buyers it is. Dust's open source posture is a genuine differentiator. AgentWorks is closed-source but ships compliance tooling (EU AI Act classification, Article 12 logs, an optional PII-detection layer) that the open-source Dust core does not. The right choice depends on whether code provenance or shipped compliance matters more to your decision-makers.
- Are AgentWorks and Dust really competitors? Both are European.
- Yes — both target the same buyer (EU mid-market teams that need governed AI without US-vendor friction). The Dutch / French headquarters do not change that. Healthy competition between EU vendors is exactly what the EU AI Act ecosystem needs, and we publish balanced comparisons because customers ask.
- Which platform has stronger workspace integrations?
- Roughly equivalent for the standard set (Slack, Notion, GitHub, Drive, Confluence). Dust has Salesforce + Intercom natively. AgentWorks supports custom integrations via MCP, which lets you connect any internal API without waiting for a connector.
- Does AgentWorks have French-language support?
- Models are language-agnostic (GPT-5, Claude, Mistral all support French to native level). UI is currently English; French UI is on the 2026 roadmap. Customer support is available in English and Dutch today; French support is rolling out.
- How does pricing compare for a 25-person team?
- Dust Pro is €29/seat/month, so a 25-person team pays roughly €725/month whether or not everyone uses it. AgentWorks charges a flat plan price plus token usage at published per-model rates, so the bill follows actual use: a team where everyone is in AI all day can land in a similar range, while a team with a core group of daily users generally pays less. The honest way to compare is to take one month of your own expected volume and price it against the published rates.