Claude Opus 5 is shown hitting 449.46× speedup on a kernel benchmark
scaling01 · x · 2026-07-25
A comparison table in the image shows Claude Opus 5 hitting 449.46× speedup on the kernel task, ahead of Claude Mythos 5 (430.93×) and Claude Opus 4.7 (371.75×).
The same table also reports:
- Time series forecasting: Opus 5 scored 5.68 MSE on the hard variant, versus 4.78 for Opus 4.7 and 4.51 for Mythos 5.
- LLM training (easy): Opus 5 averaged 68.54× speedup, compared with 50.67× for Opus 4.7 and 69.61× for Mythos 5.
- The threshold column translates speedups into human-effort equivalents, e.g. 4× = 1 hour, 200× = 8 hours, 300× = 40 hours.
More from Models
- OCR is not solved: SOTA model reads one 1881 line, outputs 16,421 dots — vanstriendaniel · 2026-10-26
- Gemini 4 Argon Lands 6th on RSI-Inference Bench, First Non-OA/Anthropic Model to Pass Day-0 Tasks — DynamicWebPaige · 2026-10-26
- Android Codex bug misreports model tiers: Astra Max shows as Astra Medium — SIGKITTEN · 2026-10-26
- Zvi digests Gemini 4 Argon cards: 217-page safety report, 0.7% injection success — TheZvi · 2026-10-26
- Nemotron 3.5 Super VL tested on 300 real bills: reasoning lifts accuracy 91% to 98% — Arindam_1729 · 2026-10-26
- Nanbeige4.3-33B benchmarked: 59 tok/s single stream, ~1,060 tok/s across 64 concurrent runs — MaziyarPanahi · 2026-10-26