AnswersPublished12 sources
Doubao Work Leads SuperCLUE’s Office-Agent Benchmark
Doubao Work ranked first in SuperCLUE’s October 2026 office agent benchmark with 85.33, ahead of Kimi Work at 84.57 and third place WorkBuddy. The benchmark assessed finished Word, PowerPoint, Excel and chart deliverables across document processing, materials drafting and data analysis.
Published byEdited with GPT-6 LunaImages generated with GPT Image 2
Research answer
Key takeaways
- Doubao Work ranked first in SuperCLUE’s October 2026 office agent benchmark with 85.33, ahead of Kimi Work at 84.57 and third place WorkBuddy.
- The benchmark assessed finished Word, PowerPoint, Excel and chart deliverables across document processing, materials drafting and data analysis.
- The available results confirm the top three and the benchmark’s scope, but not a complete eight product score table or detailed category scores.
AI Prompt
openai.comCreate a landscape editorial hero image for this Studio Global article: What did SuperCLUE’s October 2026 SuperCLUE-Office end-to-end office-agent benchmark, covered in a November 5 report citing SuperCLUE’s publ. Article summary: SuperCLUE’s October 2026 office-agent benchmark placed Doubao Work first with 85.33 overall, followed by Kimi Work and Tencent’s WorkBuddy in second and third place. The available excerpts confirm those headline results . Topic tags: general, general web. Style: premium digital editorial illustration, source-backed research mood, clean composition, high detail, modern web publication hero. Use reference image context only for broad subject, composition, and topical grounding; do not copy the exact image. Avoid: logos, brand marks, copyrighted characters, real person likenesses, fake screenshots, UI text, readable text, watermarks, charts with fake numbers, clic
ByteDance’s Doubao Work took first place in SuperCLUE’s October 2026 office-agent benchmark, scoring 85.33 overall. Moonshot AI’s Kimi Work came second with 84.57, and Tencent’s WorkBuddy ranked third. The results point to a close contest at the top, but they should be read as a leaderboard for this particular test—not as a verdict on every office workflow. 4
17
What the benchmark tested
SuperCLUE evaluated eight domestic office-agent products on end-to-end office work. The test covered document processing, materials drafting and data analysis, divided into 13 secondary tasks. It focused on the quality of finished Word, PowerPoint, Excel and chart deliverables, rather than conversational responses alone. 17
The benchmark also assessed five capabilities: file and environment operations; planning and carrying out work across multiple files; following constraints; content quality and professionalism; and respecting safety and permission boundaries. 17
The confirmed leaderboard results
| Rank | Product | Overall score |
|---|---|---|
| 1 | Doubao Work | 85.33 |
| 2 | Kimi Work | 84.57 |
| 3 | Tencent WorkBuddy | Not specified in the available report excerpts |
The supplied results identify the remaining participants as Baidu Dazi, QwenWork, SenseTime’s Raccoon Work, iFLYTEK Spark Work and 360’s Nano Work, but do not provide enough detail here to verify their individual ranks or scores. 17
Doubao Work’s lead over Kimi Work was 0.76 points, calculated from the reported overall scores. That narrow difference suggests the top two were close on the benchmark’s aggregate measure; it does not establish how they compare on any one task or on a particular company’s files. 4
What the results do—and don’t—tell office teams
Because the evaluation emphasized completed office files and included file handling, constraints and permissions, it offers a more workflow-oriented signal than a test of text answers alone. Still, an overall score can’t show whether an agent will handle a team’s templates, data, formatting rules or access boundaries well. Those details are best checked against the work the team actually needs done.
The available reporting confirms the top-three order and broad evaluation design, but doesn’t substantiate the complete eight-product scoreboard or detailed category scores and gaps. It therefore doesn’t support firm claims about which agent led a specific category, or about the reported spread between the top four and the highest- and lowest-scoring products.
The practical takeaway is to treat SuperCLUE’s results as one useful comparison under a defined set of tasks. For buyers, the benchmark can help identify products to test further; choosing an office agent still depends on how it performs on real documents, spreadsheets, drafting assignments and permission requirements.
Studio Global AI
Continue your research
This page includes a source-backed answer you can continue inside Studio Global.
People also ask
What is the short answer to "Doubao Work Leads SuperCLUE’s Office-Agent Benchmark"?
Doubao Work ranked first in SuperCLUE’s October 2026 office agent benchmark with 85.33, ahead of Kimi Work at 84.57 and third place WorkBuddy.
What are the key points to validate first?
Doubao Work ranked first in SuperCLUE’s October 2026 office agent benchmark with 85.33, ahead of Kimi Work at 84.57 and third place WorkBuddy. The benchmark assessed finished Word, PowerPoint, Excel and chart deliverables across document processing, materials drafting and data analysis.
What should I do next in practice?
The available results confirm the top three and the benchmark’s scope, but not a complete eight product score table or detailed category scores.