AgentRealTaskBench
中文 Request a test

Round 01

Scorecard

Untested means blank. We never guess a score.

Round 01 progress: 0 of 600 runs done
AgentOverallSafetyCoverageOnline tasksTravelRecommendationsPurchasingEmailProactive helpRoutinesIntegrationsPrivacyMemoryPersonalityPhone callsGroupsChained tasksRestraintContent creationData accessLocal apps: China
AsmiAsmi——0/20——————————————————
ChatGPT agentOpenAI——0/20——————————————————
ClaudeAnthropic——0/20——————————————————
DotsOpenAI——0/20——————————————————
DoubaoByteDance——0/20——————————————————
GeminiGoogle——0/20——————————————————
Grok BotxAI——0/20——————————————————
Hermes AgentNous Research——0/20——————————————————
InstinctInstinct——0/20——————————————————
KimiMoonshot AI——0/20——————————————————
ManusManus——0/20——————————————————
MuseMeta——0/20——————————————————
OpenClawOpen source——0/20——————————————————
PallyPally——0/20——————————————————
PerplexityPerplexity——0/20——————————————————
PokePoke——0/20——————————————————
sznszn——0/20——————————————————
TabTab——0/20——————————————————
TodayToday——0/20——————————————————
TomoTomo——0/20——————————————————
  1. AsmiAsmiOverall—Safety—Coverage0/20
  2. ChatGPT agentOpenAIOverall—Safety—Coverage0/20
  3. ClaudeAnthropicOverall—Safety—Coverage0/20
  4. DotsOpenAIOverall—Safety—Coverage0/20
  5. DoubaoByteDanceOverall—Safety—Coverage0/20
  6. GeminiGoogleOverall—Safety—Coverage0/20
  7. Grok BotxAIOverall—Safety—Coverage0/20
  8. Hermes AgentNous ResearchOverall—Safety—Coverage0/20
  9. InstinctInstinctOverall—Safety—Coverage0/20
  10. KimiMoonshot AIOverall—Safety—Coverage0/20
  11. ManusManusOverall—Safety—Coverage0/20
  12. MuseMetaOverall—Safety—Coverage0/20
  13. OpenClawOpen sourceOverall—Safety—Coverage0/20
  14. PallyPallyOverall—Safety—Coverage0/20
  15. PerplexityPerplexityOverall—Safety—Coverage0/20
  16. PokePokeOverall—Safety—Coverage0/20
  17. sznsznOverall—Safety—Coverage0/20
  18. TabTabOverall—Safety—Coverage0/20
  19. TodayTodayOverall—Safety—Coverage0/20
  20. TomoTomoOverall—Safety—Coverage0/20