Chinese AI model slashes prices while closing the gap with US rivals

The gap between China’s leading AI models and top US competitors has narrowed to a single-digit margin, with the newest release from Moonshot AI scoring 55 points on the Artificial Analysis Intelligence Index. This places Kimi K3.1 Pro in fourth position globally, just three points behind the current leader, Claude Opus 5.5. For readers in New York and New Jersey watching the tech sector closely, this marks a significant shift in the competitive landscape, as a Chinese model now sits within striking distance of the proprietary leaders from Anthropic and other major players.

The performance jump is substantial, representing an 11-point increase over its predecessor, Kimi K3. Only three models currently rank higher: Claude Opus 5.5 at 58 points, followed by Claude Sonnet 5.5 and Muse Wonder Full, both at 56. If Moonshot follows through on its plan to release the model weights by November 4, 2026, K3.1 Pro would become the highest-scoring open-weights model available. This move would shrink the performance gap between open and proprietary models from 12 points to just three, assuming no other major releases occur in the interim.

The improvements are most visible in agentic tasks and coding. On the GDPval-AA benchmark, the new model achieves 1,846 Elo, a 312-point leap from the previous version, placing it just behind Claude Opus 5.5 and Muse Wonder Full. In terminal-based testing, K3.1 Pro reports a 65.7 percent score on Terminal-Bench 4.0, slightly trailing Claude Opus 5.5’s 66.4 percent but well ahead of GPT-6 Astra. Hallucination rates have also dropped sharply, with claim-level errors falling from 23.4 percent to 8.3 percent on Moonshot’s internal stress tests, though GPT-6 Astra still holds an edge with a 5.5 percent rate.

Beyond logic and code, the model shows strong capabilities in 3D modeling and finance. It can interact directly with game engines like Unity and Unreal Engine, scoring 92.7 percent on the BenchCAD benchmark when using tools. In financial modeling, it gained 7.2 points over its predecessor, and it currently tops all tested models on the OfficeQA Pro v2 benchmark for general office tasks. Moonshot has also launched a "3D Make" feature that allows users to preview and export 3D models directly via its platform.

Despite the performance gains, the model maintains the same architectural size as its predecessor, with 2.8 trillion total parameters and 104 billion active ones. However, the cost of using the model has decreased by 33 percent. Input tokens now cost $2 per million, down from $3, while output tokens are $10, reduced from $15. This price drop makes the model more competitive for developers in high-cost markets like New York and New Jersey, where API expenses can quickly add up for enterprise applications.

Moonshot has also addressed cybersecurity concerns by stripping cyber-related data from the training set for the open-weights release. This deliberate limitation reduces the model’s ability to solve offensive cyber tasks from 90.5 percent to 81.0 percent on internal benchmarks, a trade-off the company argues is necessary to balance offensive and defensive capabilities. A faster variant, K3.1 Pro Highspeed, is also available, running two to three times faster than the standard version.

The stakes for the AI industry are high as this model approaches the frontier of current capabilities. By offering near-top-tier performance at a lower price point, Moonshot is pressuring US competitors to justify their premium pricing. For businesses in the Northeast corridor, the availability of a highly capable, cost-effective model from a Chinese provider adds a new option to the stack, potentially influencing adoption rates and driving further price competition in the coming months.