Compare top models

LLM Benchmarks

See how today's leading models write SpacetimeDB code, scored task by task on live modules.

DeepSeek V4 Pro

Typescript · With Guidelines
97.8%Task Pass Rate
87/89Checks Passed
47Total Evals

Pass Rate Over Time

Click a data point to view results from that date.

auth

100% · 8 evals
StatusTask IDChecksRun Time
t_026_auth_identity_check1/121.4s
t_027_private_vs_public_table1/114.6s
t_041_registered_user_gate1/116.9s
t_042_admin_bootstrap1/117.9s
t_043_role_based_access1/129.1s
t_044_ban_list1/123.6s
t_045_rate_limit1/125.7s
t_046_shared_document1/120.8s

basics

97% · 15 evals
StatusTask IDChecksRun Time
t_000_empty_reducers1/113.8s
t_001_basic_tables1/115.2s
t_002_scheduled_table1/116.7s
t_003_struct_in_table1/115.1s
t_004_insert2/213.6s
t_005_update3/322.7s
t_006_delete3/312.4s
t_007_crud4/413.7s
t_008_index_lookup3/319.5s
t_009_init4/410.4s
t_010_connect1/115.1s
t_011_helper_function3/317.7s
t_038_schedule_at_time1/125.4s
t_039_cancel_schedule0/129.6s
t_040_lifecycle_player1/110.9s

data modeling

100% · 6 evals
StatusTask IDChecksRun Time
t_030_two_table_join1/125.6s
t_031_unique_constraint1/115.4s
t_024_event_table1/115.6s
t_025_optional_fields1/122.5s
t_028_cascade_delete1/120.3s
t_029_filter_and_aggregate1/121.9s

queries

88% · 8 evals
StatusTask IDChecksRun Time
t_032_range_query1/115.5s
t_022_view_basic0/113.0s
t_023_view_per_user1/113.6s
t_033_sort_and_limit1/124.4s
t_034_find_first1/115.6s
t_035_select_distinct1/116.5s
t_036_count_without_collect1/121.5s
t_037_multi_column_filter1/121.0s

schema

100% · 10 evals
StatusTask IDChecksRun Time
t_012_spacetime_product_type3/315.7s
t_013_spacetime_sum_type3/320.1s
t_014_elementary_columns3/315.7s
t_015_product_type_columns3/315.9s
t_016_sum_type_columns3/311.9s
t_017_scheduled_columns2/217.5s
t_018_constraints3/315.0s
t_019_many_to_many5/520.3s
t_020_ecs7/749.5s
t_021_multi_column_index5/511.1s