Claude Opus 5 ranks just below Sol 6 and Fable 5 on LiveBench, but real-world tests lag
bindureddy · x · 2026-07-25
Bindu Reddy says Claude Opus 5 sits just below Sol 6 and Fable 5 on LiveBench, and that Anthropic appears to have “benchmaxxed” it.
He argues that initial real-world testing still puts Opus 5 below Fable 5 and Sol 6, especially in long-running loops that benchmarks miss.
Key claims
- LiveBench places Opus 5 just behind Sol 6 and Fable 5.
- In his tests, it performs below those models in real-world use.
- He says current benchmarks fail to capture long-horizon execution well.
- He also ranks Fable 5, Kimi K3, and Grok 4.5 as the top models in their price class.
Related event: Claude Opus 5 Accused of Benchmark Gaming, Lags Behind in Real Tests(2 posts)→
More from Models
- 'Doomer takeover = AI winter': Terra 6.1 flops on Terminal-Bench 4.0 vs Sonnet 5.5 — mark_k · 2026-10-27
- Together AI sits down with Meta to explain Muse Spark Open — togethercompute · 2026-10-27
- GPT-6.1 Terra hits ~60% on Agents' Last Exam at $1/task, 9x cheaper than Opus 5.5 — ChrisGPT · 2026-10-27
- ChatGPT keeps telling tech-support users to email 404 Media's Jason Koebler — petrusenko_max · 2026-10-27
- GPT-6.1 Terra called the awkward middle tier, priced at half of GPT-6.1 Sol — cedric_chee · 2026-10-27
- Mystery model "Terra" scores 48 on Artificial Analysis index with no public API — haider1 · 2026-10-27