Capabilities
Capability matrix
Documented and observed capabilities across 30 profiled products and 11 capability domains. A cell shows the strongest status among that domain's records; hover for the breakdown, click for the evidence.
Read this as coverage, not quality. “Documented” means a vendor says it exists under stated conditions. It does not mean the agent completes the task reliably — for that, see leaderboards. Outlined cells mark domains with at least one independent observation.
| Product | Autonomy & background work | Execution environments | Calls, messages & identity | Connectors & tools | Memory & personalization | Channels & interfaces | Research & deliverables | Transactions & bookings | Collaboration & parallelism | Reliability & recovery | Permissions, security & data |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Meta Muse | |||||||||||
| OpenAI dots | |||||||||||
| Grok Bot | |||||||||||
| Claude (including Cowork execution) | |||||||||||
| Manus 2.0 / Manus Studio | |||||||||||
| Lindy(Personal Lindy + Teammate) | |||||||||||
| Gemini Spark | |||||||||||
| Perplexity Computer (Cloud) | |||||||||||
| Perplexity Personal Computer (local access) | |||||||||||
| Microsoft Copilot Cowork for personal accounts(Preview) | |||||||||||
| Instinct | |||||||||||
| Town | |||||||||||
| Pally | |||||||||||
| Ollie | |||||||||||
| szn | |||||||||||
| Shuffle | |||||||||||
| Tomo | |||||||||||
| Asaply | |||||||||||
| Caddy | |||||||||||
| Cue by Manus | |||||||||||
| OpenClaw | |||||||||||
| Hermes Agent | |||||||||||
| AutoGLM (hosted application) | |||||||||||
| Doubao Phone Assistant (consumer version) | |||||||||||
| Qwen App (task execution / work assistant) | |||||||||||
| Coze 3.0 / Coze Agent (formerly Coze Space) | |||||||||||
| HONOR YOYO Agent | |||||||||||
| MiniMax Agent/MiniMax Code Work Mode | |||||||||||
| QwenWork | |||||||||||
| WorkBuddy (Tencent) |
Independently observedDocumentedPartly disclosed/testedLimited / conditionalPlannedHistorical / unconfirmedNot supportedNot disclosedUnknownNo record yetIncludes independent evidence
Detailed observations: 34 capabilities
A deeper, cell-by-cell review for three flagship persistent agents. Each cell records scope, conditions and sources; unknown cells are explicitly marked as not yet researched.
Independently observedDocumentedPartly disclosed/testedLimited / conditionalPlannedHistorical / unconfirmedNot supportedNot disclosedUnknownNo record yetIncludes independent evidence
Status vocabulary
Research files use 22 raw availability labels; PAEval displays them as 9 statuses. Counts are over all 539 records.
| Status | Meaning | Records |
|---|---|---|
| Independently observed | A third party observed or tested the behaviour. | 7 |
| Documented | The vendor documents the capability or condition. | 234 |
| Partly disclosed / partly tested | Some of the scope is documented or tested; the rest is not. | 31 |
| Limited or conditional | Available only in some regions, plans, betas, waitlists or demos. | 191 |
| Planned | Announced or in development; not available. | 1 |
| Historical or unconfirmed | Older observation, or a media report not yet confirmed by the vendor. | 8 |
| Not supported | A source explicitly states it is not supported. | 1 |
| Not disclosed | Sources were reviewed; the vendor does not say. | 18 |
| Unknown | Not established from the reviewed sources. Unknown is not unsupported. | 48 |