DeepSeek V4 Pro benchmarks close to Opus 5 on KernelBench-Hard
teortaxesTex · x · 2026-08-21
Benchmark results show DeepSeek V4 Pro achieving 9.35% of roofline performance on the TopK task in KernelBench-Hard, closely trailing Opus 5 at 9.46%. The data also compares performance across models like Kimi K3, Fable 5, and Qwen 3.8 Max on various tasks including FP8, KDA, and Paged attention.
More from Models
- Model at capacity for a straight week drives frustrated dev back to Opus — joonasvirtanen · 2026-10-16
- Decision models match LLMs on routing and safety — and are 7-8x faster — randal_olson · 2026-10-16
- Xiaomi MiMo V2.6 Pro takes #1 open-model spot on Agent Arena at $0.11/task — arena · 2026-10-16
- ChatGPT's 'best AI video tool' answer points to Runway integration, sparking ad doubts — tlakomy · 2026-10-16
- Microsoft's MAI finishes in-house 3T-parameter pretrained base models, post-training results coming — soumitrashukla9 · 2026-10-16
- Hands-on with Gemini 4 Argon: faster than Opus 5.5 but hallucinates more and costs more — prasenx · 2026-10-16