|
The agent led all three quality dimensions and placed first on 12 of 16 metrics in a seven-agent evaluation by Physion Labs.
SAN FRANCISCO, July 31, 2026 /PRNewswire/ — invideo Agent One has ranked first overall in Physion-Arc 1.0, an independent benchmark from Physion Labs that evaluates AI video agents on minute-long, multi-scene video generation. Agent One scored 72.4 out of 100, ahead of other leading Agents. It led all three quality dimensions, scoring 77.6 in Narrative Coherence, 63.5 in Cinematic Language and 79.2 in Production Quality, and placed first on 12 of 16 metrics, never falling below second on any.
Physion Labs builds evaluation systems for generative video, which it describes as world critics for world models. The lab is led by CEO Evelyn Qin Zhang, previously at AWS AI Labs with an MIT PhD in physics-based simulation and machine learning, and CTO Bing Shuai, formerly a principal scientist at AWS AI. Physion-Arc 1.0 is its third benchmark this year.
Each agent received a complete screenplay and planned, generated and assembled the video in a single end to end run. Seven agents produced 700 videos across 100 prompts, scored by human annotators on 16 metrics, with disagreements resolved by a third-party adjudicator and preferences aggregated through a Bradley-Terry model.
Agent One placed first on both metric families, scoring 82.3 on the eight objective metrics and 62.5 on the eight human preference metrics. Its strongest result was Narrative Alignment at 97.8, the highest single metric score in the evaluation. It posted the tightest spread in the field, with a standard deviation of 8.2 across 100 prompts and only three prompts below 50. Its 2.8-point lead over second place is statistically significant, with a p-value of 0.046.
Physion updated its benchmark Physion-Arc 1.0 on July 28, 2026. Physion ran it end to end, with no invideo involvement in annotation, adjudication or scoring.
“Agentic systems are the future of generative media, and of how creative work gets done. Ranking first on an evaluation we did not design is a testament to what this team has built, and it settles something worth saying plainly: this category will not be won on funding size. It will be won on technical depth and on strategy. Invideo builds for serious creatives, and a third of our own team are working creatives, which is why we built Agent One to hold the film together while the director makes the film” said invideo founder and CEO Sanket Shah.
ข่าวที่เกี่ยวข้อง
- Ascletis Announces Initiation of Phase I Study in U.S. for Oral Amylin Receptor Peptide Agonist ASC36 for the Treatment of Obesity
- Semiflor Participates in 2026 SDG Asia, Showcasing Healthy and Sustainable Practices from Materials to Spaces
- Extraordinary in Hand, Effortless in Control HUAWEI Mate XT 2 | ULTIMATE DESIGN Debuts as the First Smartphone to Feature LogicFolding Tau (τ) Chip and Reach Mass Commercialization
- Acer Reports Revenues for August at NT$30.18 Billion and Year-to-August at NT$214.81 Billion; Both Marking 13-Year Highs for the Same Period
