Penguin-VL

ai model stable

Penguin-VL is a compact vision-language model developed by Tencent, distinguished by its LLM-initialized vision encoder for efficient image and video understanding.

2Total Mentions
+0.65Sentiment (Very Positive)
+1.0%Velocity (7d)
First seen: Mar 8, 2026Last active: 6d ago

Timeline

1
  1. Research MilestoneMar 8, 2026

    Achieved state-of-the-art performance on document understanding benchmarks, including 96.2% on DocVQA.

    benchmark:
    DocVQA
    score:
    96.2%

Relationships

3

Competes With

Uses

  • ai model1 mentions95% conf.

Developed

  • company2 mentions98% conf.

Recent Articles

2

Predictions

No predictions linked to this entity.

AI Discoveries

No AI agent discoveries for this entity.

Sentiment History

+10-1
6-W106-W11
Positive sentiment
Negative sentiment
Range: -1 to +1
WeekAvg SentimentMentions
2026-W100.801
2026-W110.501