DeepSeek · Last 7 days
DeepSeek V4 Flash
0–100 · higher means more positive community experience, not benchmark performance.
Updated 2026-09-15 11:02 UTC · 2026-09-08 – 2026-09-15 UTCAs of 2026-09-15 11:27:03 UTC, DeepSeek V4 Flash in the DeepSeek family has 170 explicitly attributed community comments in the last 7 days; its community score is 73.4/100. Available category scores include General text 70.3/100 (n=62); Coding 60.1/100 (n=10); Reasoning 70.3/100 (n=14). Scoring method and sources
Explore long-term changes & analysis
Model lifecycle
DeepSeek V4 Flash
Released 2026-04-24 · Back to DeepSeek
Exact-version signals observed 2026-08-06–2026-09-14 · n=1,638
Data window: 2026-08-18 00:00:00 to 2026-09-15 00:00:00 UTC · Scoring method: experience_score_v2.
Lifecycle updated daily · latest complete UTC day 2026-09-14
- Fixed 14-day baseline
- 67.8 n=528
- Latest 28 days
- 68.7 n=837
- Change vs baseline
- +0.9
- 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.
3 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: Reasoning, Local deploy, Coding.
| Category | Baseline → current | Category change | Weight share | Experience contribution | Discussion-mix contribution |
|---|---|---|---|---|---|
| Reasoning | Baseline → current78.0 → 64.3n=74 → 81 | Category change−13.8 | Weight share15.2% → 9.9% | Experience contribution−1.73 | Discussion-mix contribution−0.20 |
| Local deploy | Baseline → current64.3 → 61.7n=35 → 44 | Category change−2.6 | Weight share5.9% → 5.1% | Experience contribution−0.14 | Discussion-mix contribution+0.04 |
| Coding | Baseline → current69.9 → 69.8n=60 → 103 | Category change−0.0 | Weight share11.5% → 12.8% | Experience contribution−0.00 | Discussion-mix contribution+0.03 |
| General text | Baseline → current63.7 → 64.4n=179 → 295 | Category change+0.7 | Weight share32.5% → 33.5% | Experience contribution+0.24 | Discussion-mix contribution−0.04 |
| Speed & latency | Baseline → current67.8 → 73.0n=157 → 297 | Category change+5.3 | Weight share30.3% → 36.6% | Experience contribution+1.75 | Discussion-mix contribution+0.19 |
| Image / vision | Baseline → currentnot enough datan=0 → 3 | Category change– | Weight share– | Experience contribution– | Discussion-mix contribution– |
| Roleplay / creative | Baseline → currentnot enough datan=9 → 7 | Category change– | Weight share– | Experience contribution– | Discussion-mix contribution– |
| Safety & refusals | Baseline → currentnot enough datan=15 → 9 | 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: +0.74 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.
Community reviews
Selected comments from the last 7 days. The balance of excerpts does not represent the share of positive reviews.
26 selected excerpts
Show original
但是现在用标准模式，烂的要死，实在是没有办法用。
Show original
个人体感就是做视觉方面的还是很差，一个任务做了快1小时，没有任何反应，用Cursor 几分钟做好了
Show original
还是v4更适合角色扮演
Show original
两台dgx跑V4flash，能有70-80token/s，个人使用足够了
No selected comments for this filter in the last 7 days.