GPT-6 Sol/Luna benchmarks leak: near-Fable performance at 80-85% lower cost
kimmonismus · x · 2026-09-23
kimmonismus shares GPT-6 Sol & Luna benchmark numbers:
- FrontierCode: Sol 48.4% vs Fable 5.1's 48.7% (both xhigh), $1.37 vs $9.27 per task (85% cheaper)
- DeepSWE: 68.8% at max vs Fable 5's best 69.9% at xhigh, $2.74 vs $13.41 (80% cheaper)
- AutomationBench: 33.2% at xhigh beats Fable 5.1 + Opus 5 fallback's 31.4%, $0.27 vs $2.45+ per task
Bottom line: GPT-6 series matches near-frontier performance at dramatically lower cost.
More from Models
- Google's Physical-Engineering's Last Exam: all 14 frontier models score below 50% — dl_weekly · 2026-11-12
- Commit leak reveals unreleased Gemini 4 variants: Argon, Hydrogen and Helix (unconfirmed) — BlackHC · 2026-11-11
- Rewiring just 89 of 442,368 neurons in Qwen3-8B matches full SFT for a persona — maxsloef · 2026-11-11
- Unsloth ships Qwen3.8-27B Dynamic V3.1 GGUFs with 5-10% accuracy gains — danielhanchen · 2026-11-11
- Zvi: being forced back to an older model is the best reminder of AI progress — TheZvi · 2026-11-11
- Screenshots Made a 4B Browser Agent Worse: A Postmortem With Real Numbers — Stunning-Sherbet1853 · 2026-11-11