AI ranking
Business agents
Agents aimed at small businesses and business automation (prospecting, no-code, sales assistants). Assessed on business outcome, supervision and compliance.
Visual top 10 — Le Recul score
- 01 ChatGPT Agent 76.50
- 02 Microsoft Copilot Cowork 76.50
- 03 Gemini Enterprise 73.80
- 04 Claude Cowork 62.10
- 05 Genspark 61.20
- 06 Manus 60.30
- 07 Microsoft Copilot 58.50
- 08 Google Workspace Studio 58.50
Compare models. Tick up to 3 models in the table below, then run the comparison: who wins on which use, by how much, and how solid the measurement is.
| # | Model | Vendor | Score | Assessment | |
|---|---|---|---|---|---|
| 01 | OpenAI agent built into ChatGPT, able to browse the web and handle files. | OpenAI | 76.50 / 100 | Le Recul | |
| 02 | Cowork variant of Microsoft Copilot for multi-agent collaboration. | Microsoft | 76.50 / 100 | Le Recul | |
| 03 | Google enterprise offering built on Gemini, with customisable agents. | 73.80 / 100 | Le Recul | ||
| 04 | Anthropic collaborative workspace built on Claude — email, docs, research, analysis. | Anthropic | 62.10 / 100 | Le Recul | |
| 05 | Web research and synthesis agent, an alternative to Perplexity. | Genspark ai | 61.20 / 100 | Le Recul | |
| 06 | Autonomous multi-task agent positioned as a general-purpose alternative to cloud assistants. | Manus ai | 60.30 / 100 | Le Recul | |
| 07 | Microsoft agentic suite built into Microsoft 365, Outlook and Teams. | Microsoft | 58.50 / 100 | Le Recul | |
| 08 | Google agentic studio built into Workspace (Gmail, Docs, Sheets). | 58.50 / 100 | Le Recul |
Indicative score, weighted from public sources. Transparent methodology. See the AI ranking page for the full methodology and the other categories.
AI agent ranking for small business
Le Recul's Business agents ranking covers agents aimed at small and medium businesses: prospecting, no-code automation, sales and administrative assistants. They are assessed on the business outcome obtained, not on the sales demonstration.
This is the segment where the gap between promise and reality is widest. These offerings are sold on time saved and extra revenue, rarely on the cost of setting them up, the supervision required or the acceptable error rate. Yet that is precisely what decides whether an agent holds up over time.
Public coverage of this category is thin: few independent measurements, a great deal of vendor communication. The ranking therefore stays deliberately cautious and reads as a reference point, not as a buying recommendation.
What this ranking is for
This ranking helps you assess a business agent on what actually matters to a small company: time genuinely saved, reliability of the workflow, full cost and the level of supervision required.
It also helps spot the blind spots: compliance, handling of customer data and liability in the event of an error remain yours, whatever the tool.
How to read the ranking
- Real, measurable business outcome, not the sales promise.
- Time actually saved, net of the time spent checking.
- Reliability of the workflow over time.
- Available integrations and ease of setting it up.
- Full cost and compliance risk, customer data included.
FAQ — Business agents
Can an AI agent really replace a sales task?
It can absorb part of one — sorting, follow-ups, preparation — rarely the whole thing. Tasks that commit the customer relationship need human validation.
Why is this ranking so cautious?
Because independent measurements are rare here. Showing confident scores based on data largely supplied by vendors would be misleading.
What is the main hidden cost?
Supervision. An agent that handles ninety per cent of cases correctly still requires checking, and it is that time that is usually missing from the return-on-investment calculation.
Who is liable if the agent makes a mistake?
The company using it, in almost every case. Vendors' terms of use nearly always limit their own liability.
How often is the ranking updated?
About every 25 hours, automatically. The update date is shown at the top of the page.