Four AI Giants Unleash Frontier Models in Record Time

In just three weeks, four top AI labs released frontier language models, marking a significant acceleration in the development of advanced artificial intelligence. On July 8, xAI launched Grok 4.5, which can run at lower cost and faster than Opus 4.8 from Anthropic. The model was trained on real developer sessions from Cursor, a coding platform acquired by SpaceXAI earlier this year.

Grok 4.5's pricing is $2 per million input tokens and $6 per million output tokens, with a 500,000 token context window. Elon Musk described the model as an Opus-class model that runs faster and at lower cost. Independent trackers place it fourth on the Artificial Analysis Intelligence Index, ahead of every open weight model.

On July 24, Anthropic released Claude Opus 5, which is now the default model on Claude Max subscriptions. The model ships with a 1 million token context window and an adjustable reasoning effort setting that ranges from low to high mode. Opus 5 sets new marks on Frontier-Bench and GDPval-AA, two coding and knowledge work evaluations.

OpenAI also released GPT-5.6 family models, which ship as three tiers: Sol, Terra, and Luna. The Sol tier is priced at $5 input and $30 output per million tokens and leads the Artificial Analysis Coding Agent Index with a 91.9% score on Terminal-Bench 2.1.

Moonshot AI released Kimi K3 on July 16, a 2.8 trillion parameter mixture of experts model that carries a 1 million token context window and native multimodal input. The model is priced at $3 input and $15 output per million tokens and has been committed to publishing full open weights by July 27.