ModelBrief · Kimi K3.1 Pro · Topic Page
Last updated Nov 20, 2026

Kimi K3.1 Pro vs Kimi K3: What Post-Training Added

ModelBrief · Published 2026-11-16 · Fact-check 2026-11-16 · Reference: Moonshot tech blog, Artificial Analysis, LMArena, Agent Arena

Comparing Kimi K3.1 Pro with Kimi K3 is a rare controlled experiment: the same 2.8T base, the same parameter count, with a single architectural change — 24 Gated DSA layers replacing full-attention Gated MLA layers — and everything else coming from substantially larger post-training compute. The score gap between the generations is the net return on post-training — large, but very unevenly distributed.

What changed: three things

First, one architectural change: 24 Gated DeepSeek Sparse Attention (Gated DSA) layers replace K3's full-attention Gated MLA layers; the other 69 Kimi Delta Attention layers are unchanged, and total parameters stay 2.8T with 104B active. Second, substantially larger post-training compute, with joint RL across hundreds of domains. Third, the commercial terms changed: K3 carried a Modified MIT license, while K3.1 Pro ships under the custom Kimi K3.1-Pro License (a separate agreement for MaaS revenue over $20M; products with over 100M MAU or over $20M monthly revenue must display the model name). The first two are technical upgrades; the third is a "regression" — from a standard open license to a custom license with commercial conditions.

The AA Index: 44 to 55, the biggest jump on an independent board

Artificial Analysis Intelligence Index v4.3.2: K3 scored 44, K3.1 Pro 55, No. 4 overall. An 11-point jump is post-training's hardest independent evidence — same base, same evaluation set, so the gap can only come from post-training. One caution: on the v5.0 re-version K3.1 Pro scores 46, and K3's v5.0 score was not verified in this research round, so 46 must not be compared with 44 across versions. Agent Arena corroborates: net improvement K3 +0.9%, K3.1 Pro +5.1% (Max); Bash Reliability K3 −4.5%, K3.1 Pro +7.9%. Long-horizon coding: Kimi Code Bench 3.0 (internal), K3 55.95%, K3.1 Pro 61.2%.

Finance and hallucination: where the post-training returns concentrate

The vendor's internal finance benchmarks (self-reported): Modeling 50.4 → 57.6, Research 58.8 → 70.4, Business Operation 8.9 → 37 (~4.2×). OfficeQA Pro v2 (finance) 64.4, SOTA per Moonshot; Frontier Finance 58.6. Hallucination rate (Kimi Hallu Bench Hard, claim-level): with tools 23.4% → 8.3%; without tools K3.1 Pro is 5.4% (GPT-6 Astra 4.2%, Opus 5.5 6.8%). These two sets of numbers show where the post-training money went: finance tasks and factual reliability — exactly the agent and office scenarios Kimi targets.

Internal finance benchmarks, two generations (vendor-reported) Kimi K3.1 Pro Kimi K3 Modeling 57.6 50.4 Research 70.4 58.8 Business Operation 37 8.9 Moonshot internal finance benchmarks (points; vendor-reported); orange is Kimi K3.1 Pro
Internal finance benchmarks (points; Moonshot tech blog, vendor-reported; orange is Kimi K3.1 Pro, grey is Kimi K3)

What didn't change: LMArena leads by only 0.6 points

LMArena Text: K3.1 Pro 1488.0, No. 24; K3 1487.4 — a 0.6-point gap, inside the ±13.0 confidence interval, not significant. On general chat, post-training opened almost no gap. The contrast is worth remembering: AA Index +11, finance Business Operation ~4.2×, hallucination down by nearly two-thirds — all task-specific returns; the general-chat return is near zero. Post-training made it directionally stronger, not uniformly stronger.

Cyber: 33.3% to 81.0% — but 81.0% is the weakened figure

xdev-bench-hard pass@3: K3 33.3%, the released K3.1 Pro 81.0% — on the surface a 2.4× jump. But 81.0% is the deliberately weakened figure: Moonshot removed cyber-related training data, and the uncleaned checkpoint scored 90.5%. The real post-training gain (33.3% → 90.5%) is larger than the released number shows, while the 81.0% users actually get is the vendor's downgraded version. Both numbers must be stated together in a generational comparison, or the post-training gain will be understated and the released capability overstated.

Price: stronger, yet about a third cheaper

K3.1 Pro's standard tier is $2/$10, about a third less than Kimi K3 cost at launch; AA's cost per index task is $2.11, basically level with K3's $2.00. A stronger model at a lower price — uncommon in generational upgrades, and explained by post-training costing less at the margin than training a new base. Transition arrangement: Kimi K3 (kimi-k3) is served by K3.1 Pro until December 7, 2026; whether the K3 model ID continues after that date was not verified in this research round.

FAQ

Should I upgrade from K3 to K3.1 Pro?

For agent and finance tasks: yes — AA Index +11, Business Operation ~4.2×, hallucination down by nearly two-thirds. For general chat: the gain is near zero, with LMArena up only 0.6 points, not significant. Also note the license change: K3's Modified MIT became a custom license with commercial conditions.

Is K3.1 Pro a brand-new model?

No. It is a post-trained product of Kimi K3: the same 2.8T base, with the only architectural change being 24 Gated DSA layers replacing full-attention Gated MLA layers; everything else comes from post-training compute.

Can I still use K3?

Kimi K3 (kimi-k3) is served by K3.1 Pro until December 7, 2026; arrangements after that date were not verified in this research round.

ModelBrief's take: K3 to K3.1 Pro is a textbook post-training upgrade: same base, AA +11, large finance and hallucination gains, a lower price. The only cost is the license — from Modified MIT to a custom license. For users the upgrade decision is simple: upgrade for agent and finance work, don't bother for general chat; for the open community, the license change matters more than the scores.