DeepSeek V4 Flash 0731 (Reasoning, Max Effort)

deepseek-v4-flash-0731 · 2 samples · combined / raw

Scoring

Choose the quality benchmark and final score formula for this model.

Cost adjustment

Tune formulas that account for Cost per Task.

Score over time

54.454.654.955.255.423:0023:1523:3023:4500:00 Run #1394 · Score · X: 2026-08-03 23:00:05 · Y: 54.9Run #1395 · Score · X: 2026-08-04 00:00:05 · Y: 54.9

Cost per task over time

$0.0171$0.0221$0.0271$0.0321$0.037123:0023:1523:3023:4500:00 Run #1394 · Cost per task · X: 2026-08-03 23:00:05 · Y: $0.0271Run #1395 · Cost per task · X: 2026-08-04 00:00:05 · Y: $0.0271

Time per task over time

5.69m5.72m5.75m5.78m5.81m23:0023:1523:3023:4500:00 Run #1394 · Time per task · X: 2026-08-03 23:00:05 · Y: 5.75mRun #1395 · Time per task · X: 2026-08-04 00:00:05 · Y: 5.75m
Run?Fetched?Score?Quality?Cost/task?Time/task?Intel?Code?Agent?MMMU%?
#1395 2026-08-04 00:00:05 54.9 54.9 $0.0271 5.75m 49.9 69.1 45.7 -
#1394 2026-08-03 23:00:05 54.9 54.9 $0.0271 5.75m 49.9 69.1 45.7 -