Moonshot AI released Kimi K3.1 Pro on October 23, 2026, and it immediately landed fourth on the Artificial Analysis Intelligence Index with a score of 55 — just three points behind the top-ranked model. Why it matters: a Chinese model that Moonshot plans to release as open weights is now within striking distance of the best closed models in the world, and it got cheaper at the same time.
October 23, 2026
What happened
K3.1 Pro is a post-trained version of Moonshot’s Kimi K3, built on the same base model with significantly more post-training compute, according to the company. Its Intelligence Index score of 55 is an 11-point jump over K3’s 44. Only three models rank higher: Anthropic’s Claude Opus 5.5 at 58, Claude Sonnet 5.5 at 56, and Meta’s Muse Wonder Full at 56, according to Artificial Analysis. The score puts K3.1 Pro ahead of OpenAI’s GPT-6 Astra and Google’s Gemini 4 Argon, both at 53.
Moonshot says it plans to release the model’s weights by November 4, 2026. If that happens on schedule and the rankings hold, K3.1 Pro would become the highest-scoring open-weights model available, narrowing the gap between open and proprietary models on the index from 12 points to 3.
The details
The biggest gains are in agentic work. On GDPval-AA, which measures performance on real-world work tasks, K3.1 Pro reached 1,846 Elo — 312 points above K3 and just behind Claude Opus 5.5 (1,866) and Muse Wonder Full (1,857), according to Artificial Analysis. On AutomationBench-AA, its score rose from 58% with K3 to about 73%.
The model keeps the same dimensions as K3: 2.8 trillion total parameters with 104 billion active, a one-million-token context window, and text and image input. Moonshot also shipped a variant called K3.1 Pro Highspeed that the company says runs two to three times faster. Both versions are available now through Kimi.ai, Kimi Work, Kimi Code, and the Kimi API.
Prices fell even as performance rose. API token prices dropped 33% across the board: input is now $2 per million tokens (down from $3), output is $10 (down from $15), and cache hits are $0.20 (down from $0.30). The Highspeed variant costs $4 per million input tokens and $20 per million output tokens. Because the new model generates more output tokens per task, Artificial Analysis puts its cost per Intelligence Index task at $1.98, almost identical to K3’s $2.00.
Moonshot also says it stripped cyber-related data from both the mid-training and reinforcement-learning stages before the planned open-weights release, stating that no open-weights model should extend the absolute frontier of offensive cyber capability. On the company’s internal Hallu Bench Hard stress test, it reports claim-level hallucinations fell from 23.4% with K3 to 8.3% — a company-reported figure from an internal test, not an independent measurement.
Why it matters
For anyone choosing a model today, the practical takeaway is price and access. A top-four model at $2/$10 per million tokens, with a faster variant and open weights promised within two weeks, puts real pressure on closed frontier pricing — especially for agentic coding and office-style work, where K3.1 Pro’s gains are concentrated. The honest caveat is verification: several of the headline benchmark gains come from Moonshot’s own testing at maximum reasoning effort, so independent hands-on results will matter more than the launch table. That is the everything-AI-tested view: the ranking from Artificial Analysis is independent and genuinely impressive; the rest should be treated as claims to check once the weights land on November 4.
Frequently asked questions
What is Kimi K3.1 Pro?
It is Moonshot AI’s newest reasoning model, released October 23, 2026. It is a post-trained upgrade of Kimi K3 with the same 2.8 trillion total parameters, a one-million-token context window, and text and image input.
How does Kimi K3.1 Pro rank against Claude and GPT models?
It scores 55 on the Artificial Analysis Intelligence Index, ranking fourth — behind Claude Opus 5.5 (58), Claude Sonnet 5.5 (56), and Muse Wonder Full (56), and ahead of GPT-6 Astra and Gemini 4 Argon (both 53).
When will Kimi K3.1 Pro’s open weights be released?
Moonshot says it plans to release the weights by November 4, 2026. Until then, the model is available through Kimi.ai, Kimi Work, Kimi Code, and the Kimi API.
How much does Kimi K3.1 Pro cost via API?
$2 per million input tokens, $10 per million output tokens, and $0.20 per million cache-hit tokens — a 33% cut from Kimi K3’s prices. The Highspeed variant costs $4 input and $20 output per million tokens.
Sources: The Decoder, OfficeChai, Artificial Analysis (via The Decoder and OfficeChai reporting)