As of 9 Oct 2026, PAEval tracks 6 products similar to OpenAI dots (persistent personal agent from OpenAI). Similarity is based on product type, product group, shared independent evaluators and documented capabilities. The order reflects similarity, not quality, and PAEval does not recommend one product over another.
Meta · United States · Persistent personal agent
2026-10-09 Consumer/Small Business Muse; fixed Free/Power/Maximum, region, cloud or Mac connection status; not equivalent to Muse Spark model, Muse Code or the meditation app of the same name.
Independent scores: 9.3/10 on Assistant Benchmark; 20% trusted completion on micro1 PersonalAgentBench (n=10)
Google · United States · Persistent personal agent
Spark personal execution mode within Gemini Apps; Beta/experimental nature. Cloud tasks, local Chrome, Mac file access and packages should be fixed separately. Gemini 3.5 and Antigravity harness were released in May; 3.7 Flash was clearly upgraded on August 13th. Flash has entered the Gemini app in September 3.8, but this time there is no equivalent clear update to Spark’s current orchestration model.
Independent scores: 60% trusted completion on micro1 PersonalAgentBench (n=10)
xAI (via Cursor) · United States · Individual/Team Agent
Current desktop/mobile Grok Bot; visible changelog v0.68.0 (10-05), unfixed background model; personal Bot, Team Bot and template classification, recording Cursor/SuperGrok access and Auto Review. Not Grok Chat, Grok Build, or the Model API.
Independent scores: 7.3/10 on Assistant Benchmark; 30% trusted completion on micro1 PersonalAgentBench (n=10)
Anthropic · United States · General-purpose work agent
On 2026-09-16, it was announced that chat and Cowork were merged into Claude, and Pro/Max was gradually launched; Team/Enterprise still needs to distinguish between the old selector and organization switch. Claude Code is not incorporated into the bank's ability score.
Independent scores: none published yet
Manus · Singapore · General-purpose work agent
2.0 was released on 2026-09-28, using the Cascade execution framework; the desktop workspace was upgraded to Manus Studio. Cue is another independent application that does not integrate its wallet, phone calls, and agent group chat into the main Manus product. Product independence does not mean data independence: the help page states that Manus/Cue share accounts and data, and deleting the account affects both; Cue phone subscriptions and weekly quotas are independent of Manus points, and skill configurations must also be set separately.
Independent scores: none published yet
Lindy · United States · Individual/Team Agent
The current official website/document is positioned as individual and team AI teammate: private Lindy manages emails/meetings, and shares Slack teammates to process team reports/CRM. There is no verification of public semantic version numbers. The old visual agent builder/telephone sales product capabilities cannot be automatically merged into the current Teammate.
Independent scores: none published yet