Compare top models

LLM Benchmarks

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

qwen/qwen3-coder-flash

Typescript · With Guidelines
57.3%Task Pass Rate
51/89Checks Passed
47Total Evals

Pass Rate Over Time

Click a data point to view results from that date.

auth

75% · 8 evals
StatusTask IDChecksRun Time
t_026_auth_identity_check1/14.3s
t_027_private_vs_public_table1/15.0s
t_043_role_based_access1/16.0s
t_041_registered_user_gate1/15.4s
t_042_admin_bootstrap1/15.3s
t_044_ban_list1/15.7s
t_045_rate_limit0/15.1s
t_046_shared_document0/16.3s

basics

83% · 15 evals
StatusTask IDChecksRun Time
t_000_empty_reducers1/15.8s
t_001_basic_tables1/15.0s
t_002_scheduled_table1/14.1s
t_003_struct_in_table1/13.0s
t_004_insert2/24.2s
t_005_update3/36.4s
t_006_delete3/35.1s
t_007_crud0/45.4s
t_008_index_lookup3/36.2s
t_009_init4/45.2s
t_010_connect1/14.9s
t_011_helper_function3/33.8s
t_038_schedule_at_time1/16.3s
t_039_cancel_schedule0/16.2s
t_040_lifecycle_player1/15.3s

data modeling

83% · 6 evals
StatusTask IDChecksRun Time
t_025_optional_fields1/15.5s
t_024_event_table1/15.3s
t_028_cascade_delete1/15.3s
t_029_filter_and_aggregate1/17.2s
t_030_two_table_join0/16.8s
t_031_unique_constraint1/14.9s

queries

50% · 8 evals
StatusTask IDChecksRun Time
t_022_view_basic0/15.3s
t_023_view_per_user1/14.5s
t_032_range_query1/15.7s
t_033_sort_and_limit1/16.7s
t_034_find_first0/13.9s
t_035_select_distinct0/14.4s
t_036_count_without_collect0/15.7s
t_037_multi_column_filter1/16.1s

schema

30% · 10 evals
StatusTask IDChecksRun Time
t_012_spacetime_product_type3/34.5s
t_013_spacetime_sum_type3/36.3s
t_014_elementary_columns3/34.0s
t_015_product_type_columns0/35.9s
t_016_sum_type_columns0/36.8s
t_017_scheduled_columns2/27.5s
t_018_constraints0/34.7s
t_019_many_to_many0/57.9s
t_020_ecs0/77.9s
t_021_multi_column_index0/57.5s