agents benchmark
VIBE-V2
VIBE-V2 is an internal benchmark covering pure front-end and full-stack Web, Android, and iOS projects with build-from-scratch tasks. It uses an Agent-as-a-Verifier paradigm to automatically verify program interaction logic and visual output, scoring models through a unified pipeline that includes a requirement set, containerized deployment, and a dynamic interaction environment.
Updated Aug 11, 2026
Models1
Model coverage1
MetricScore
EvidenceB
VIBE-V2 Ranking
Higher score ranks better on this benchmark.
| 1 | MI | 50.1% | 100.0% | 1 | C |
VIBE-V2 Highlights
The leading models and scores on this benchmark.
What is VIBE-V2?
What VIBE-V2 measures and how its scores work.
VIBE-V2 is an internal benchmark covering pure front-end and full-stack Web, Android, and iOS projects with build-from-scratch tasks. It uses an Agent-as-a-Verifier paradigm to automatically verify program interaction logic and visual output, scoring models through a unified pipeline that includes a requirement set, containerized deployment, and a dynamic interaction environment.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- VIBE-V2
- Modality
- text
- Primary category
- agents
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- vibe-v2|llm-stats-current
Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about VIBE-V2.
Which model scores highest on VIBE-V2?
MiniMax M3 is currently ranked first with 50.1%.
What does VIBE-V2 measure?
VIBE-V2 is an internal benchmark covering pure front-end and full-stack Web, Android, and iOS projects with build-from-scratch tasks. It uses an Agent-as-a-Verifier paradigm to automatically verify program interaction logic and visual output, scoring models through a unified pipeline that includes a requirement set, containerized deployment, and a dynamic interaction environment.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
1 model results are currently shown.
Does this benchmark affect the overall score?
No. This benchmark is shown for reference but does not contribute to the overall score.