Frontier Models See Intensive Updates in 8 Days
ArtificialAnlys · x · 2026-07-18
Artificial Analysis noted a flurry of frontier model updates within 8 days: Grok 4.5, GPT-5.6, Muse Spark 1.1, and Kimi K3 dropping successively.
Key takeaways:
- 6 labs now have models scoring over 50 on the Artificial Analysis Intelligence Index, a significant jump from just 2 in early June.
- Kimi K3 entered the overall leaderboard at No. 3 with a score of 57. It excels in agentic and knowledge work capabilities, rivaling or beating some more expensive models across multiple benchmarks.
- Inference costs for near-frontier models dropped 2–3x in a single week. The article highlights that per-task costs for GPT-5.6, Grok 4.5, Muse Spark 1.1, and Kimi K3 are significantly lower than peer models from just days ago.
- Claude Fable 5 remains No. 1, but its lead is shrinking, and the price of "underlying intelligence" continues to drop.
Related event: Artificial Analysis Says Four Frontier Models Launched in Eight Days(5 posts)→
More from Infra
- Encrypted GPT-2 speedrun challenges FHE inference: best is 3.031 s/token, goal sub-1s — tensorqt · 2026-10-27
- Open-source Latch compiles PyTorch into training megakernels, 2.75x faster on H100 — sguada · 2026-10-27
- Stanford launches GPU-centric supercomputer Marlowe for data science and AI research — StanfordHAI · 2026-10-27
- NVIDIA Claims HGX Rubin NVL8 Delivers 10x More Tokens per Megawatt Than B200 — thursdai_pod · 2026-10-27
- Together lets multiple LoRA post-training experiments share the same GPU capacity — togethercompute · 2026-10-27
- Commoditization dilemma: either human labor or GPUs become the commodity — _AustinCalvert_ · 2026-10-27