AI Benchmarks: Kimi K3.1 Pro Narrows Gap With Top Rivals

Chinese artificial intelligence firm Moonshot AI has unveiled Kimi K3.1 Pro, a new reasoning model that has quickly climbed the global rankings. The system now holds the fourth position on the Artificial Analysis Intelligence Index with a score of 55. This represents an eleven point improvement over its predecessor, Kimi K3. The new model sits just three points behind the leading proprietary systems, marking a significant shift in the competitive landscape.

K3.1 Pro is a post-trained derivative of K3 that utilized substantially more computing power during its final training phase. According to data from Artificial Analysis, only three models currently outperform it: Anthropic's Claude Opus 5.5 with 58 points, and both Claude Sonnet 5.5 and Muse Wonder Full with 56 points each. Moonshot intends to release the model weights by November 4, 2026. If this timeline holds, K3.1 Pro will become the highest scoring open-weights model available. This move would shrink the performance gap between open and proprietary models from twelve points to just three, assuming no other major changes occur in the interim.

Alongside the standard version, the company is offering K3.1 Pro Highspeed. This variant is designed to run two to three times faster than the original. Both versions are accessible through Kimi.ai, Kimi Work, Kimi Code, and the company's API. The most notable gains in performance have occurred in agentic tasks. On the GDPval-AA benchmark, the new model reached an Elo rating of 1,846. This is 312 points higher than K3 and places it just behind Claude Opus 5.5 and Muse Wonder Full.

Coding capabilities have also seen substantial improvement. On Terminal-Bench 4.0, Moonshot reports a score of 65.7 percent. This result exceeds GPT-6 Astra, which scored 58.2 percent, and trails only Claude Opus 5.5 at 66.4 percent. By comparison, the previous Kimi K3 managed only 13 percent on that specific test. The company also highlights a sharp reduction in hallucinations. On its internal Kimi Hallu Bench Hard, claim-level hallucinations dropped from 23.4 percent with K3 to 8.3 percent. While GPT-6 Astra still leads this metric at 5.5 percent, Moonshot notes that its benchmark serves as a stress test rather than a measure of general error rates.

The model's architecture remains consistent with K3, featuring 2.8 trillion total parameters with 104 billion active. It supports a one-million-token context window for text and images. Artificial Analysis positions it at the top of the Pareto frontier, indicating that no model of similar size achieves a higher score. API prices have decreased by 33 percent, with input costs at $2 per million tokens and output at $10. Cache hits are priced at $0.20. Although the model generates 14 percent more output tokens, the cost per Intelligence Index task remains nearly identical to K3 at $1.98.

Moonshot has also emphasized advancements in 3D modeling and cybersecurity. The system works with game engines like Unity, Unreal Engine, and Godot. It can write and run scripts for Blender and CAD software without requiring MCP connectors. On the BenchCAD test, it scores 92.7 percent with tools, slightly behind Claude Opus 5.5 and GPT-6 Astra. For its open-weights release, the company removed cyber-related data from the training process to balance offensive and defensive capabilities. This adjustment reduced its score on an internal vulnerability reproduction test from 90.5 percent to 81.0 percent, though it still outperforms several other open models.