Model lifecycle
Kimi K3
Released 2026-07-16 · Back to Kimi
Exact-version signals observed 2026-08-07–2026-09-03 · n=481
Data window: 2026-08-07 00:00:00 to 2026-09-04 00:00:00 UTC · Scoring method: experience_score_v2.
Lifecycle updated daily · latest complete UTC day 2026-09-03
- Fixed 14-day baseline
- 60.6 n=156
- Latest 28 days
- 62.7 n=481
- Change vs baseline
- +2.1
- 90-day slope
- – 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.
2 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.
Largest negative experience contributions: Speed & latency.
| Category | Baseline → current | Category change | Weight share | Experience contribution | Discussion-mix contribution |
|---|---|---|---|---|---|
| Speed & latency | Baseline → current34.9 → 34.1n=30 → 95 | Category change−0.8 | Weight share21.4% → 20.7% | Experience contribution−0.16 | Discussion-mix contribution+0.18 |
| Coding | Baseline → current63.6 → 68.5n=20 → 60 | Category change+5.0 | Weight share13.4% → 13.1% | Experience contribution+0.65 | Discussion-mix contribution−0.02 |
| Reasoning | Baseline → current80.3 → 86.4n=36 → 97 | Category change+6.1 | Weight share24.1% → 20.8% | Experience contribution+1.36 | Discussion-mix contribution−0.77 |
| General text | Baseline → current57.6 → 65.2n=53 → 189 | Category change+7.7 | Weight share31.4% → 37.7% | Experience contribution+2.64 | Discussion-mix contribution+0.07 |
| Image / vision | Baseline → currentnot enough datan=0 → 3 | Category change– | Weight share– | Experience contribution– | Discussion-mix contribution– |
| Local deploy | Baseline → currentnot enough datan=7 → 18 | Category change– | Weight share– | Experience contribution– | Discussion-mix contribution– |
| Roleplay / creative | Baseline → currentnot enough datan=4 → 9 | Category change– | Weight share– | Experience contribution– | Discussion-mix contribution– |
| Safety & refusals | Baseline → currentnot enough datan=6 → 10 | Category change– | Weight share– | Experience contribution– | Discussion-mix contribution– |
| Video generation | Baseline → currentnot enough datan=0 → 0 | Category change– | Weight share– | Experience contribution– | Discussion-mix contribution– |
Unresolved contribution from categories below the sample threshold: −1.86 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.