Domains: Coding
Skills: CodingMultimodalInstruction FollowingReasoning

Four-Factor Integrity Breakdown & Primary-Source Evidence

Privacy

Weight 40%
7.5/10
Access Model: Hybrid Public / Private
"We evaluate 114 of the 168 briefs in Code2Video Bench: 39 from HeyGen's public split and 75 private briefs shared with us by HeyGen."

Originality

Weight 25%
10/10
Reuse Profile: All Original (0% reuse)
"Each model gets one brief and writes one HyperFrames HTML composition: HTML for layout and GSAP for animation. We render the composition to a video with HyperFrames."

Freshness

Weight 20%
9/10
Initial Release: 2026-09Latest Refresh: 2026-10Rolling Updates: No
"Data as at 26 October 2026 (dataset 0.3.0)"

Grading

Weight 15%
10/10
Integrity: Independent Offline (Leak-Verified)Sandbox: offline isolated
"We render each composition to an MP4 video with HyperFrames 0.8.30, with no network access and a 5-minute limit."

Frontier Model Standings on Code2Video-AA

Top 15 active models tracked
Rank Model Official Score Percentile
#1Claude Opus 5.5 max Anthropic1536 Elo88.5p
#2Claude Sonnet 5.5 max Anthropic1443 Elo84.9p
#3GPT-6 Astra max OpenAI1427 Elo81.2p
#4GPT-6.1 Sol max OpenAI1369 Elo77.6p
#5Claude Fable 5.1 max Anthropic1325 Elo73.9p
#6Gemini 4 Argon high Google1259 Elo70.2p
#7Claude Haiku 5.5 max Anthropic1192 Elo66.5p
#8Grok 4.7 xhigh SpaceXAI1127 Elo62.9p
#9DeepSeek V4.1 Flash max DeepSeek1053 Elo59.2p
#10MiMo-V2.6-Flash default Xiaomi1052 Elo55.6p
#11MiMo-V2.6-Pro default Xiaomi1052 Elo51.9p
#12GLM-5.3 max Z AI1034 Elo48.2p
#13GPT-6 Luna max OpenAI1030 Elo44.6p
#14Qwen3.8 Max (0902) xhigh Alibaba1021 Elo40.9p
#15Inkling 2 Small default Thinking Machines1020 Elo37.3p

Embeddable Audit Badge

View Markdown (.md)

Benchmark maintainers can embed this live SVG badge in GitHub READMEs or documentation:

BenchAudit Code2Video-AA Tier S 8.8/10
[![BenchAudit Tier S](https://benchaudit.org/badges/code2video-aa.svg)](https://benchaudit.org/benchmarks/code2video-aa/)