Cities　experience　the　most　intensive　level　of　human　activity.　As　a　result,　more　than　60　%　of　global　CO2　emissions　come　from　urban　areas.　Urban　green　space　has　the　dual　ecological　benefits　of　increasing　carbon　sinks　and　reducing　carbon　emissions.　Creating　green　space　is　essential　to　promoting　the　development　of　a　low-carbon　cycle　in　a　city.　Therefore,　exploring　the　quantitative　structure　and　spatial　pattern　optimization　of　urban　green　space　from　the　perspective　of　carbon　balance　can　effectively　improve　the　total　carbon　sink　of　a　city.　Based　on　the　carbon　balance　theory,　this　paper　first　evaluates　the　carbon　offsetting　capability　(COC)　of　urban　green　space　in　Beijing　in　2020.　Then,　CO2　emissions　is　predicted,　COC　improvement　targets　are　established,　and　the　quantity　of　standard　green　space　is　calculated　under　these　targets　in　2035.　A　multi-objective　programming　model　(MOP)　is　constructed　to　derive　the　optimal　solution　to　determine　the　amount　of　standard　green　space　needed　to　meet　the　constraints　of　urban　development　planning　and　maximize　the　carbon　sink.　A　circuit　model　is　used　to　identify　the　priority　distribution　area　of　green　space　in　2035,　and　the　Patch-generating　Land　Use　Simulation　(PLUS)　model　is　used　to　simulate　the　spatial　pattern　optimization　results.　The　results　show　that:　(1)　CO2　emissions　in　Beijing　caused　by　human　activities　in　2020　totaled　about　240.12　million　tons,　the　net　absorption　of　CO2　of　green　space　was　about　8.99　million　tons,　and　the　COC　was　about　3.74　%;　(2)　in　2035,　Beijing＇s　CO2　emissions　will　be　about　265.40　million　tons.　Under　5　%,　10　%,　15　%,　20　%,　and　25　%　COC　improvement　gradients,　the　demand　for　standard　green　space　will　be　12,204.80　km2,　12,763.80　km2,　13,353.85　km2,　13,943.90　km2,　and　14,533.96　km2,　respectively;　(3)　the　results　of　the　Multi-Objective　Programming　(MOP)　model　show　that　the　optimal　COC　in　2035　is　4.19　%,　and　the　standard　green　space　quantity　is　13,012.24　km2;　(4)　In　2035,　Beijing＇s　urban　green　space　will　be　characterized　by　a　network　structure　of　circular　radiation.　The　optimized　COC　is　4.37　%,　and　the　standard　green　space　quantity　is　13,577.86　km2,　which　is　largely　consistent　with　the　MOP　model＇s　predicted　result.　This　study　can　provide　theoretical　and　methodological　support　for　urban　green　space　planning,　and　expand　nature　-based　solutions　for　the　low-carbon　development　in　cities.