Moonshot AI: Kimi K3.1 Pro ranks fourth on intelligence index
The new model scores 55 points, placing it three points behind top rivals and cutting API prices by 33 percent.
By Sarah Montgomery ·
A Chinese artificial intelligence company has released a reasoning model that now places fourth on a major global intelligence ranking. Moonshot AI introduced Kimi K3.1 Pro on October 23, 2026. The system scored 55 points on the Artificial Analysis Intelligence Index.
The score represents an 11-point increase over its predecessor, Kimi K3. Only three proprietary systems currently hold higher positions on the index. Anthropic’s Claude Opus 5.5 leads with 58 points. Claude Sonnet 5.5 and Muse Wonder Full both score 56 points. Moonshot plans to publish the model weights by November 4, 2026. If released, K3.1 Pro would become the highest-scoring open-weights model available. That move would reduce the gap between open and proprietary models from 12 points to 3.
The company reports substantial gains in agentic tasks. On the GDPval-AA benchmark, K3.1 Pro reached 1,846 Elo. This figure is 312 points above K3 and trails only Claude Opus 5.5 and Muse Wonder Full. On Terminal-Bench 4.0, Moonshot reports a score of 65.7 percent. This result exceeds GPT-6 Astra’s 58.2 percent but falls slightly short of Claude Opus 5.5’s 66.4 percent. The previous version, K3, managed only 13 percent on that test.
Performance improvements extend to specific professional domains. Moonshot states that claim-level hallucinations dropped from 23.4 percent in K3 to 8.3 percent in K3.1 Pro on its internal Kimi Hallu Bench Hard. The company notes that GPT-6 Astra still leads this metric at 5.5 percent. In finance, the model gained 7.2 points on Financial Modeling and 11.6 points on Financial Research. Moonshot says it tops all tested models on OfficeQA Pro v2.
The model maintains the same architecture as K3, featuring 2.8 trillion total parameters and 104 billion active parameters. It supports a one-million-token context window and handles both text and images. Artificial Analysis places K3.1 Pro on the Pareto frontier, indicating no model of comparable size scores higher on the index. API prices dropped by 33 percent. Input costs now stand at $2 per million tokens, with output at $10 and cache hits at $0.20.
Moonshot also detailed cybersecurity capabilities in its technical blog. The company states that vulnerability research improved as post-training scaled up. For the open-weights release, it removed cyber-related data from both training phases. On its internal xdev-bench-hard, the uncleaned model solved 90.5 percent of tasks involving disclosed vulnerabilities in software such as the Linux kernel and Node.js. After cleaning, the score stood at 81.0 percent. DeepSeek-V4.1-Flash scored 85.7 percent, while Kimi K3 scored 33.3 percent.
Both the standard and Highspeed variants of the model are available through Kimi.ai, Kimi Work, Kimi Code, and the Kimi API. The Highspeed variant runs two to three times faster according to the company. Users can access new 3D modeling features on Kimi.com, which allow for previewing, editing, and exporting models without MCP connectors.
