# Kimi K3.1 Pro: the same 2.8T model, more RL, and what the −65% counts

> Satyajit Ghana — Head of Engineering @ Inkers Technology
> canonical: https://ai.thesatyajit.com/articles/kimi-k3-1-pro
> date: 2026-10-24
> tags: kimi, mixture-of-experts, reinforcement-learning, benchmarks, agentic-coding, pricing

Moonshot launched **Kimi K3.1 Pro** at 17:15 UTC on 23 October, after a teaser earlier the same day. The
[launch post](https://x.com/Kimi_Moonshot/status/2113680701770068067) calls it "our most capable 3T-class open-weight
model": a 1M-token context, 2.8T total and 104B active parameters, multimodal, full-project game builds in Unreal,
Unity and Godot, parametric 3D in Blender and CAD, finance hallucination "down by 65%", and "matching the strongest
proprietary models across coding, agentic and knowledge benchmarks". It is live on the Kimi API now, with a
**Highspeed** variant "2–3× faster". The weights are promised **by 4 November 2026**.

They are not out yet. On 24 October the Hugging Face API lists one K3 repository under `moonshotai`, the
July `Kimi-K3`, and nothing for K3.1 (measured). So there is no `config.json` to diff, no tokenizer to read and no
safetensors header to count. Everything below about the model itself is what Moonshot says, labelled that way. What I
can check is the launch blog's own benchmark table, the pricing pages, a third-party measurement, and the arithmetic
of the demo that went viral with it.

If you have not read the [Kimi K3 article](/articles/kimi-k3), the short version: K3 is a 93-layer MoE that routes
each token to 16 of 896 experts, mixes Kimi Delta Attention with Gated MLA 3:1, reaches back across depth with
Attention Residuals, and was post-trained as nine RL experts merged into one model by on-policy distillation. None of
that changes here.

## What changed from K3

The blog's first technical sentence settles the architecture question: "Kimi K3.1 Pro is a post-trained version of
Kimi K3, with substantially larger post-training compute" (reported). Same base, same shape. The 2.8T / 104B, the
1,048,576-token window and the text, image and video inputs in the launch post are K3's numbers, and the API docs say
"All K3 series models support a 1,048,576-token context window." OpenRouter lists both K3.1 Pro endpoints with the
quantization tag `mxfp4`, which matches the MXFP4 expert weights K3 was trained to serve in (third-party listing).

So the diff is not in the weights' shape. It is in training data, RL scale, price and speed:

<K3Diff />

The one row that does not fit "post-trained version" is the cyber row. The blog says that for the open-weight release
Moonshot "identified and removed cyber-related data from both training stages": the mid-training corpus and the RL
environment mix. You cannot remove data from a stage you did not rerun, so a mid-training stage was redone for K3.1
Pro, not only RL stacked on the K3 checkpoint (reasoned). That is still "the same base model" in the sense that
pre-training was not redone, and it is the same move Z.ai made with [GLM-5.3](/articles/glm-5-3), which kept
GLM-5.2's base and spent a month on post-training.

For reference, here is the architecture the K3.1 Pro weights should carry, from K3's own report. If the 4 November
`config.json` differs from K3's, this is the line that will be wrong:

<Figure
  src="https://media.thesatyajit.com/ai/articles/kimi-k3/fig3.png?v=3b8379515c5a1b44"
  alt="The Kimi K3 architecture. Right: a stack of blocks, each containing three KDA layers and one Gated MLA layer, every attention layer paired with a Stable LatentMoE feed-forward network, with alpha-gated Attention Residual connections reaching back to the embedding and all preceding block outputs. Top left: the Stable LatentMoE module with shared and routed experts behind a router. Bottom left: the KDA module. Bottom right: the native vision pathway, MoonViT-V2 into an MLP projector."
  caption="The Kimi K3 architecture, which Moonshot says K3.1 Pro shares unchanged (Kimi K3 tech report, Figure 2)."
/>

## Where the gain comes from: scaling RL, not the model

If the network is the same, every point K3.1 Pro gains over K3 was bought with post-training. Moonshot's account of
how is one paragraph, and it is about plumbing, not a new objective. To run "large-scale joint RL across hundreds of
downstream domains" they optimized "the RL framework, long-context inference, training efficiency, sandbox
infrastructure, rollout scheduling, and a shared platform for environment quality checks", so that "each increment of
RL compute delivers stable, predictable returns" (reported).

That list reads like a list of the ways RL spending gets wasted, and it is worth taking one item at a time:

- **Long-context inference** is most of the cost. An agent's RL step is mostly generation: long trajectories against
  tools before any gradient. K3 already paused straggler trajectories and resumed them next iteration so a few long
  ones would not stall the batch.
- **Sandboxes and rollout scheduling** decide how many environments are live at once. K3's AgentENV microVMs sit
  idle up to 98% of their lifetime waiting on the model, and pausing them is what made 51 million sandboxes affordable.
- **Environment quality checks** are the quiet one. Hundreds of domains means hundreds of reward functions, and one
  that pays for the wrong behaviour poisons the shared policy. K3's GPU-kernel reward already needed a growing list of
  hacking detectors.
- **"Joint"** is the change from K3, whose RL trained nine separate experts, three domains by three effort levels,
  and merged them with Multi-Teacher On-Policy Distillation. Joint RL across hundreds of domains is a different
  recipe, but the blog does not say whether the expert-then-distil structure survived.

The blog gives no number for the compute: no GPU-hours, no ratio to K3's post-training, no curve of score against
RL steps behind "stable, predictable returns", and no technical report yet.

The results that can be laid side by side with K3 are all Moonshot's (reported): Kimi Code Bench 3.0 Long-horizon from
55.95 to 61.2, financial modeling up 7.2 points, financial research up 11.6, a simulated business operation at about
4.2× K3's score, and xdev-bench-hard, Moonshot's exploit-writing benchmark, from 33.3 to 81.0. One independent number
agrees on direction: Artificial Analysis's current Intelligence Index has K3 (Max) at 43.6 and K3.1 Pro at 55.1
(third-party, read from AA's model page on 24 October; that is a newer index version than the v4.1 score in the K3
article, so the two are not comparable).

## The benchmark table, not the chart

The launch chart is 14 bar panels. The blog's table, at the bottom of the same page, has 47 rows across coding,
agentic, knowledge-and-vision and cyber, against five models: Claude Fable 5.1, GPT-6 Astra and Claude Opus 5.5 on the
proprietary side, DeepSeek-V4.1-Flash-1017 and GLM-5.3 on the open side. Here is the chart:

<Figure
  src="https://media.thesatyajit.com/ai/articles/kimi-k3-1-pro/fig1.png?v=a54e61dff60937e5"
  alt="Fourteen horizontal bar charts comparing Kimi K3.1 Pro (blue) with Opus 5.5, GPT-6 Astra, Fable 5.1, DeepSeek-V4.1-Flash-1017 and GLM-5.3 on Terminal-Bench 4.0, SWE Marathon, Terminal-Bench Science 0.1, Kimi Code Bench 3.0 Long-horizon, Agents' Last Exam, BenchCAD with tools, OfficeQA Pro v2, Frontier Finance, GDPVal-AA v2.1, OSWorld V2.1, Humanity's Last Exam Full with tools, GDP.pdf, ChartoGraphy with tools and ZeroBench with tools. Kimi K3.1 Pro is first in six panels and second in five."
  caption="The launch chart: 14 of the table's 47 rows. A note says all models were maxed out on thinking effort, Claude models may include fallbacks and GPT models may include cyberguards (Moonshot, Kimi K3.1 Pro tech blog)."
/>

I pulled the table out of the JSON the page ships and ranked every row (measured from the page). With the three
proprietary models as the comparators, 45 rows have at least one of their scores, and K3.1 Pro is ahead or tied on
**16** of them. Counting all five comparators, it is first on 16 of 47 rows, and by category the picture is lopsided:
first on 13 of 25 agentic rows, 2 of 10 coding rows, 1 of 9 knowledge-and-vision rows, and 0 of 3 cyber rows. The chart
is a mild selection: K3.1 Pro is first in 6 of its 14 panels, against 10 of the other 33 rows (measured). Program
Bench (31.5 against Opus 5.5's 65.0) and HLE-Diamond (39.4 against GPT-6 Astra's 65.8) are in the table and not the
chart.

Toggle the comparators and categories to see which claims hold against which models:

<BenchCompare />

Four things in the footnotes change how much any single row says (all reported):

1. **Who ran what.** K3.1 Pro was run by Moonshot at max effort, temperature 1.0, top-p 0.95, mostly in the **Claude
   Code** harness, but in Kimi Code for BioMysteryBench and Agents' Last Exam and in the Kimi Agent harness for
   OSWorld. Most comparator scores were copied from each lab's report or a public leaderboard, run in their own
   harnesses. Terminal-Bench 4.0, the headline row, takes "the maximum value reported in either their blog posts or
   the official Terminal-Bench 4.0 leaderboard" for everyone else.
2. **Fallbacks and guards.** Every Claude column is labelled "max, with fallback" and the GPT-6 Astra column too. The
   chart note adds that "GPT Models may include cyberguards". Both are mechanisms that tend to lower a comparator's
   score, so if anything they cut against the comparators, not for them.
3. **Zeros assigned by Moonshot.** On OfficeQA Pro the blog says some external models "retrieved reference answers
   online; Consequently, we assigned a score of zero to these tasks", and Agents' Last Exam zeroes runs that bypass the
   task. Both are defensible. Both are a lab adjudicating its competitors' transcripts.
4. **Internal rows.** Kimi Code Bench 2.0 and 3.0, Business Operation, DECK-Arena, Design Bench, both Kimi
   Hallucination rows and xdev-bench-hard are Moonshot's own. Drop them in the widget and the agentic lead
   survives; it rests mostly on public benchmarks.

So "matching the strongest proprietary models" is true of agentic work with tools: browsing, document and office
tasks, Agents' Last Exam, Automation Bench. Terminal-Bench 4.0 is the honest headline, 65.7 against Opus 5.5's 66.4
and GPT-6 Astra's 58.2, a near tie in a harness that is not Moonshot's. It is not true of closed-book knowledge, of
the harder vision sets, or of Program Bench, where K3.1 Pro scores 31.5 and Opus 5.5 65.0. Artificial Analysis's
index lands in the same place: K3.1 Pro at 55.1 sits between Opus 5.5 (Max) at 57.6 and GPT-6 Astra (Max) at 52.7
(third-party).

## What the −65% counts

"Hallucination down by 65%" comes from one sentence in the blog: on "Kimi Hallu Bench Hard", "an internal benchmark
designed to stress-test factual accuracy, its claim-level hallucination rate fell from 23.4% for K3 to 8.3%"
(reported). That is a 64.5% relative drop (reasoned), which the post rounds to 65%. The same sentence gives the
response-level number, which is less dramatic: "the share of responses containing at least one factual error was just
42.4% of its predecessor's rate", a 57.6% drop.

A claim-level rate means a model's answer is split into individual factual claims and each claim is checked; the
footnote says the checking is done by "fact-checking agents". Two things follow. The rate depends on how the splitter
counts claims, so a model that writes more small, safe claims looks better per claim than per response, which is
why the response-level 42.4% is the more conservative figure. And the checker is an LLM agent, so the benchmark
measures agreement with Moonshot's fact-checker, not with ground truth.

<Figure
  src="https://media.thesatyajit.com/ai/articles/kimi-k3-1-pro/fig2.png?v=0bb23aa6f23f28c1"
  alt="Two bar charts titled Hallucination, lower is better. Kimi Hallu Bench without tools: GPT-6 Astra 4.2, Kimi K3.1 Pro 5.4, Opus 5.5 6.8, Fable 5.1 11.4, DeepSeek-V4.1-Flash-1017 13.1, GLM-5.3 14.8. Kimi Hallu Bench Hard with tools: GPT-6 Astra 5.5, Kimi K3.1 Pro 8.3, Opus 5.5 15.0, GLM-5.3 22.6, DeepSeek-V4.1-Flash-1017 23.3, Fable 5.1 23.3."
  caption="Moonshot's hallucination benchmark. K3 is not on this chart; its 23.4% is in the blog text. GPT-6 Astra is lower on both (Moonshot, Kimi K3.1 Pro tech blog)."
/>

The blog is careful to say the rates "do not represent overall error rates in daily use". On its own chart GPT-6
Astra is lower than K3.1 Pro on both hallucination rows, 5.5 against 8.3 on the hard set.

The more interesting finding is on the API platform. The K3.1 Pro quickstart at `platform.kimi.ai`, read on 24 October,
carries what looks like an earlier draft of the blog. Its version of the same paragraph says the rate "fell from
23.35% for K3 to 6.85%", with the error-containing share at "46.4% of its predecessor's rate", and it describes the
set differently: "de-identified conversations where users had flagged factual errors in earlier models's responses"
(measured, from the page). The same page still has unfilled slots, `[xx.x%]` where the Terminal-Bench numbers go.

Two readings follow, both reasoned. The numbers moved between drafts, from a 70.7% drop to a 64.5% drop, so the
published 65% is the more cautious of the two. And if the set really is built from prompts where users flagged errors
in *earlier* models' answers, it is selected on the predecessor's failures. A model rerun on its own worst prompts
regresses toward its average, so part of K3's 23.4% is the selection itself, and K3.1 Pro's drop overstates the
improvement on a fresh sample. The published blog no longer says how the set was built, and it is internal, so I
cannot check which description is current.

## Highspeed: same price for a cache hit, double for everything else

Highspeed is the variant with the least explanation. The model list calls it "the high-speed variant of Kimi K3.1
Pro — same capabilities at about 2-3× the speed". Prices: \$4.00 per million input tokens and \$20.00 output, exactly
double the standard \$2.00 and \$10.00, with cached input unchanged at \$0.20 (reported). OpenRouter lists it as a
second endpoint of the same model, tagged `moonshotai/highspeed`, with the same `mxfp4` quantization and the same
943,718-token output cap (third-party listing). No tokens-per-second figure is published for either, and Artificial
Analysis does not list Highspeed yet. The 2–3× is unverified.

How would the same weights at the same precision go faster? The usual levers, none confirmed by Moonshot (reasoned):

- **Smaller batches per replica.** Decode for a big MoE is memory-bound: each step reads the active weights and the
  KV and KDA state once, whether it serves one sequence or many. Batching amortizes that read across users, which is
  what makes tokens cheap and each user slow. Serve fewer sequences per replica, or spread one replica over more GPUs,
  and each user's step gets shorter while each GPU's throughput falls. The
  [GB300 decode article](/articles/kimi-k3-gb300-megakernel) shows the trade on K3 itself: one user on 16 GPUs at
  batch 1 is a configuration chosen for speed, not for cost.
- **Harder speculative decoding.** K3 ships an [EAGLE-3-style draft](/articles/eagle-3-speculative-decoding) built from
  its multi-token-prediction layer. More draft tokens per step raise speed for one user and waste compute on rejected
  drafts, which also costs throughput.

Both cost GPU-time per token, which fits a doubled per-token price. The unchanged cache price fits too: a cache hit
skips prefill, and the speed-up is in decode. That is consistent; it is not evidence.

The standard endpoint already moved. Artificial Analysis measures K3.1 Pro's median output speed at 77.1 tokens per
second against 38.5 for K3 (Max), and its average time per Intelligence Index task at 609 seconds against 1,174
(third-party). That is about 2× on the regular endpoint before Highspeed enters (reasoned), and since the network is
the same, it is a serving change or shorter reasoning, not a faster model.

<Figure
  src="https://media.thesatyajit.com/ai/articles/kimi-k3-1-pro/fig4.png?v=a86e20624f4f07be"
  alt="A pricing table titled Kimi K3.1 Pro API Pricing, subtitled K3.1 Pro launches at 2/3 of K3's price. Kimi-K3.1-Pro: input 2.00 dollars per million tokens, output 10.00, cached input 0.20. Kimi-K3.1-Pro-Highspeed: input 4.00, output 20.00, cached input 0.20."
  caption="Launch pricing: K3.1 Pro at two-thirds of K3's list price, Highspeed at twice K3.1 Pro's except for cached input (Moonshot, launch thread on X)."
/>

## The Spider-Man demo, checked against list price

The demo that travelled with the launch was a [post](https://x.com/bubbubbdex/status/2113701194187059221) of a
Spider-Man city game: "Kimi K3.1 Pro built Manhattan, added web swinging between skyscrapers, then threw in street
combat. 8 hours. 3.8M tokens. Total cost: \$34.72." Three things do not add up.

**The arithmetic.** At list price, 3.8M K3.1 Pro tokens cost between \$0.76, if every token is a cache hit, and \$38.00, if
every token is output (reasoned). To land on \$34.72 with nothing cached, 3.39M of the 3.8M tokens, 89%, must be output.
On Highspeed it takes 1.22M output tokens, 32%, still with no cache hits. Neither looks like an agent session. Moonshot's
own cyber table gives the mix of a long agentic trajectory: 159.8M input and 546K output tokens per successful
xdev-bench-hard run, 0.34% output (reported). Artificial Analysis's runs of K3.1 Pro across its whole index used 5.39B
input and 148M output tokens, 2.67% output, and 94% of input was served from cache; they cost \$3,117 in total, about
\$0.56 per million tokens (third-party token counts and costs, my division). At that rate 3.8M tokens is about \$2.14, and
\$34.72 buys about 62M tokens. One reply on the post makes the same point from experience, estimating an hour of agent
work at ten to twenty million tokens. The post claims under half a million an hour.

<DemoCost />

**The clock.** The demo was posted at 18:36 UTC, 81 minutes after the 17:15 launch post (measured from the two posts'
timestamps). Eight hours of building needs access from before the announcement. Early testers exist for most launches,
so this is not impossible, but the post does not say so.

**The provenance.** The developer behind an earlier project, [spiderbench](https://github.com/xikhar/spiderbench),
posted a similar browser Spider-Man city on 27 September, made with Claude Opus 5.5 using Blender, image generation and
Three.js, and published its source and assets. He replied that the K3.1 Pro post is his project with new lighting, and
says a community note was added. I compared the two clips frame by frame at 2 fps and found no identical frames: the
repost is relit at golden hour and has no minimap (measured). That neither proves nor rules out a derived build, and I
make no claim either way. What I can say is that the post's cost and token numbers are not consistent with each other
at Moonshot's prices, so the demo is not evidence for K3.1 Pro's cost or ability. The blog's own examples, built "end to
end by K3.1 Pro", are the ones to judge it by, and none of them comes with a token count either.

## Cyber: capability removed on purpose

The cyber section is the part of the blog with the most actual numbers, and it is unusual enough to note. During
development, Moonshot saw vulnerability-research scores rise steadily with post-training scale, traced the source to
cyber content in the mid-training corpus and cyber tasks in the RL mix, and removed both for the open-weight release.

<Figure
  src="https://media.thesatyajit.com/ai/articles/kimi-k3-1-pro/fig3.png?v=63fffdd6d8178ebb"
  alt="A table titled Internal xdev-bench-hard, pass@3. Columns: harness, problems solved at pass@3, average success rate, average tokens per successful trajectory as input and output. GPT-5.6 Sol, codex: 100, 77.8, 93.5M and 336K, marked third-party evaluation. Kimi model without cyber risk-reduction, kimi-code: 90.5, 71.4, 216.4M and 499K. DeepSeek-V4.1-Flash, claude-code: 85.7, 57.1, 162.2M and 1.5M. Kimi K3.1 Pro, kimi-code: 81.0, 64.5, 159.8M and 546K. GLM-5.3, claude-code: 57.1, 34.9, 381.1M and 3.6M. Kimi K3, kimi-code: 33.3, 15.9, 115.7M and 475K."
  caption="xdev-bench-hard: the same pipeline without the cyber data removal solves 90.5% at pass@3, the released K3.1 Pro 81.0%, K3 33.3% (Moonshot, Kimi K3.1 Pro tech blog)."
/>

The same pipeline without the removal solves 90.5% at pass@3; the final K3.1 Pro solves 81.0%; K3 solved 33.3%
(reported). So the removal cost about 9.5 points of exploit capability on Moonshot's own benchmark, while general
coding gains still lifted it far past K3. On the public cyber rows, ExploitBench and ExploitGym, K3.1 Pro is fourth
of five, behind all three proprietary models, which is what Moonshot says it is aiming for: open weights that do not
"extend the absolute frontier of cyber capability". The tokens column doubles as a cost receipt for the previous
section: a successful run is around 160M tokens.

## What to watch on 4 November

When the weights land, three things are checkable that are not checkable now:

1. **The config against K3's.** If it is the same base model, `config.json` should match K3's field for field, and
   the tensor names and shapes in the safetensors headers should be identical while every shard's hash changes. That
   is the test the [DeepSeek-V4.1-Flash-1017 diff](/articles/deepseek-v4-1-flash-1017) ran.
2. **The licence.** K3 shipped under a custom Kimi K3 License. The K3.1 Pro blog does not name one.
3. **Local speed.** Highspeed is a serving configuration, so it will not ship. What ships is the same 2.8T MXFP4
   shape that the [TPU megakernel](/articles/kimi-k3-tpu-megakernel) and [kimi-k3-in-c](/articles/kimi-k3-in-c) already
   run for K3, so both should load it unchanged if the config matches.

Until then, K3's own card is the closest public record of what K3.1 Pro's weights should look like:

<ModelCard repo="moonshotai/Kimi-K3" />

## The take

K3.1 Pro is the clearest case yet of a lab shipping a new flagship without a new network. The same 2.8T shape, the
same 1M window, an implied mid-training rerun with cyber data removed, and a large, unquantified increase in RL. What that buys is
real but narrow: K3.1 Pro leads on agentic, tool-using work and finance-flavoured document tasks, nearly ties Opus 5.5 on
Terminal-Bench 4.0, and still trails the proprietary models on closed-book knowledge, harder vision sets, Program Bench
and, by design, cyber. It costs a third less than K3 and the standard endpoint runs twice as fast.

The −65% is a real number from an internal stress test, measured per claim by an LLM fact-checker, and the per-response
figure is 58%. An earlier draft of it, still live in the API docs, said 71% and described a set selected on earlier
models' failures. And the demo is not a measurement of anything: its tokens and its price cannot both be right at
Moonshot's own rates.

---

*Sources: the [Kimi K3.1 Pro tech blog](https://www.kimi.ai/blog/kimi-k3-1-pro) (benchmark table and footnotes, finance,
hallucination and cyber sections; figures 1-3 here are its charts), Moonshot's
[launch thread](https://x.com/Kimi_Moonshot/status/2113680701770068067) (figure 4), the Kimi API
[pricing page](https://platform.kimi.ai/docs/pricing/chat), model list and K3.1 Pro quickstart, OpenRouter's endpoint
listing for `moonshotai/kimi-k3.1-pro`, and [Artificial Analysis's model page](https://artificialanalysis.ai/models/kimi-k3-1-pro)
(speed, index, token counts and run cost), all read on 24 October 2026. The architecture figure is from the
[Kimi K3 technical report](https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf). Tallies are mine, computed
from the table the blog page ships; the widgets use the same data.*