Model lifecycle

DeepSeek V4 Pro

Released 2026-04-24 · Back to DeepSeek

Exact-version signals observed 2026-06-12–2026-07-31 · n=412

Stable since baseline Compared with the first sample-sufficient 14-day window after release.

Lifecycle updated daily · latest complete UTC day 2026-07-31

Fixed 14-day baseline
66.9
n=65
Latest 28 days
68.3
n=243
Change vs baseline
+1.4
90-day slope
−0.7
pts / 30 days

14-day rolling experience

Each daily point summarizes the previous 14 complete UTC days. The dashed line is the fixed release baseline; the verdict still uses independent non-overlapping 14-day segments.

14-day rolling experience from 2026-06-19 to 2026-08-01; latest score 65.1, fixed baseline 66.9.4050607080Fixed baseline 66.92026-06-05–2026-06-19 · 66.9 · n=652026-06-06–2026-06-20 · 66.5 · n=732026-06-07–2026-06-21 · 67.0 · n=742026-06-08–2026-06-22 · 68.2 · n=772026-06-09–2026-06-23 · 66.9 · n=852026-06-10–2026-06-24 · 66.3 · n=912026-06-11–2026-06-25 · 66.4 · n=972026-06-12–2026-06-26 · 65.9 · n=1012026-06-13–2026-06-27 · 64.8 · n=1012026-06-14–2026-06-28 · 62.0 · n=892026-06-15–2026-06-29 · 63.0 · n=892026-06-16–2026-06-30 · 61.5 · n=832026-06-17–2026-07-01 · 57.8 · n=842026-06-18–2026-07-02 · 61.2 · n=892026-06-19–2026-07-03 · 62.1 · n=902026-06-20–2026-07-04 · 62.0 · n=902026-06-21–2026-07-05 · 62.0 · n=912026-06-22–2026-07-06 · 61.0 · n=892026-06-23–2026-07-07 · 62.9 · n=862026-06-24–2026-07-08 · 65.2 · n=872026-06-25–2026-07-09 · 63.1 · n=852026-06-26–2026-07-10 · 64.9 · n=902026-06-27–2026-07-11 · 64.6 · n=842026-06-28–2026-07-12 · 67.8 · n=812026-06-29–2026-07-13 · 68.6 · n=842026-06-30–2026-07-14 · 67.4 · n=852026-07-01–2026-07-15 · 69.3 · n=792026-07-02–2026-07-16 · 69.9 · n=762026-07-03–2026-07-17 · 71.8 · n=802026-07-04–2026-07-18 · 72.7 · n=792026-07-05–2026-07-19 · 71.0 · n=902026-07-06–2026-07-20 · 70.4 · n=1042026-07-07–2026-07-21 · 71.1 · n=1062026-07-08–2026-07-22 · 70.4 · n=1102026-07-09–2026-07-23 · 71.8 · n=1092026-07-10–2026-07-24 · 71.8 · n=1022026-07-11–2026-07-25 · 71.3 · n=1072026-07-12–2026-07-26 · 71.0 · n=1022026-07-13–2026-07-27 · 69.2 · n=972026-07-14–2026-07-28 · 70.4 · n=972026-07-15–2026-07-29 · 67.2 · n=1032026-07-16–2026-07-30 · 65.0 · n=1042026-07-17–2026-07-31 · 63.3 · n=1182026-07-18–2026-08-01 · 65.1 · n=16406-1906-2707-0507-1307-2107-2908-01

4 of 4 independent 14-day segments ready for trend judgment.

What explains the change

Experience change and discussion-mix change are separated. These are observational contributions, not proof of cause.

No category has enough samples in both windows for a contribution conclusion.

Category Baseline → current Category change Weight share Experience contribution Discussion-mix contribution
Speed & latency 69.4 → 69.4n=12 → 44 +0.1 19.9% → 18.6% +0.01 −0.06
General text 62.4 → 64.2n=28 → 107 +1.8 40.2% → 41.7% +0.74 −0.02
Coding 63.5 → 71.1n=14 → 51 +7.6 21.9% → 21.7% +1.65 −0.00
Image / vision not enough datan=1 → 2
Local deploy not enough datan=1 → 0
Reasoning not enough datan=9 → 39
Roleplay / creative not enough datan=0 → 0
Safety & refusals not enough datan=0 → 1

Unresolved contribution from categories below the sample threshold: −0.91 pts.

Lifecycle scores use the same experience-signal weights as the main index, without sample shrinkage after n=30. They measure public user perception, not model capability or backend causes.