AI Price Cuts: Moonshot's New Model Undercuts Rivals

Developers and businesses in New England looking to integrate advanced AI tools will see a significant shift in costs and capabilities this week. Moonshot AI has launched Kimi K3.1 Pro, a new reasoning model that not only rivals the top proprietary systems but also reduces API pricing by a third. For teams building applications, this means access to high-end intelligence without the premium price tag previously required to match leaders like Anthropic's Claude.

The new model has immediately climbed the global rankings. It currently sits at number four on the Artificial Analysis Intelligence Index with a score of 55 points. This places it just three points behind Claude Opus 5.5, which leads the proprietary pack. Only Claude Sonnet 5.5 and Muse Wonder Full rank higher on the index. If Moonshot releases the model weights as planned by November 4, 2026, it will become the highest-scoring open-weights model available.

Performance gains are most visible in complex, agentic tasks. On the GDPval-AA benchmark, the new model reached 1,846 Elo, a jump of 312 points over its predecessor. It now trails only Claude Opus 5.5 and Muse Wonder Full in this specific metric. In coding and terminal operations, the model achieved 65.7 percent on Terminal-Bench 4.0. This is a dramatic improvement from the 13 percent score of the previous version and places it nearly level with the industry leader.

The company also reports a sharp drop in hallucination rates. Internal testing shows that claim-level hallucinations fell from 23.4 percent to 8.3 percent. While GPT-6 Astra still holds an edge with a 5.5 percent rate, Moonshot notes that its test is a stress test rather than a general error measure. For finance professionals, the model showed specific gains in financial modeling and research, topping all tested models on the OfficeQA Pro v2 benchmark for general office tasks.

A major focus of the release is 3D modeling capabilities. The model can work directly with game engines such as Unity, Unreal Engine, and Godot. It scored 92.7 percent on the BenchCAD benchmark when using tools, which is close to the scores of GPT-6 Astra and Claude Opus 5.5. Moonshot has also launched a 3D Make feature on its platform that allows users to preview, edit, and export 3D models. The company states the model is particularly strong at parametric modeling of mechanical parts and architectural designs.

The model retains the same architecture as its predecessor, with 2.8 trillion total parameters and 104 billion active ones. It supports a context window of one million tokens and handles both text and images. Despite the unchanged size, the cost to use the model has dropped significantly. Input prices are now $2 per million tokens, down from $3, while output costs are $10, down from $15. Cache hits cost $0.20, a reduction from $0.30.

Moonshot has also addressed cybersecurity concerns for the upcoming open-weights release. The company removed cyber-related data from its training phases to manage offensive capabilities. On an internal benchmark involving recent software vulnerabilities, the cleaned model solved 81.0 percent of tasks. This is slightly lower than the uncleaned version but remains high compared to previous models. The company aims to balance offensive potential with defensive utility for auditing legacy systems.

Both the standard and high-speed versions of the model are available now through various interfaces. The high-speed variant runs two to three times faster than the standard version. With the open weights expected by early November, the gap between open and proprietary models is set to shrink from 12 points to just three. This development continues the trend of Chinese AI firms closing the performance gap with American rivals while offering more affordable alternatives for the broader market.