Skip to content

Model vs model

Gemini 3.1 Pro (preview) vs Kimi K3

Gemini 3.1 Pro (preview) and Kimi K3 are both in llmwise. What each gets on every plan, what it reads, how it's served and what happens when its provider fails, from the catalog and the code that runs them.

Model prices and specs checked against OpenRouter's Gemini 3.1 Pro page, OpenRouter's Kimi K3 page. Updated .

Short answer

They cost the same in llmwise: up to 125 messages a month on Pro on either. Otherwise, only Gemini 3.1 Pro (preview) reads a PDF as the whole file. In our test runs, Gemini 3.1 Pro (preview) passed 49 of the 50 prompts both answered and Kimi K3 46; 5 prompts split them, most on summarization (5 to 3).

Gemini 3.1 Pro (preview) vs Kimi K3, prompt by prompt

Every prompt Gemini 3.1 Pro (preview) and Kimi K3 both answered, compared directly, their biggest differences first. One run each, through OpenRouter: a wait depends on the provider and the load that day, so a lead under 10% counts as close.

Of the 50 prompts both answered, both passed 45, only Gemini 3.1 Pro (preview) passed 4, only Kimi K3 passed 1, and neither passed 0. Gemini 3.1 Pro (preview) answered sooner on 5 of the 50 and Kimi K3 on 42; the rest were within 10% of each other. The 50 replies cost $0.5334 on Gemini 3.1 Pro (preview) and $0.1938 on Kimi K3: 2.8× less on Kimi K3.

The 5 prompts only one of Gemini 3.1 Pro (preview) and Kimi K3 passed

  • A product announcement with five rules (writing): Gemini 3.1 Pro (preview) passed and Kimi K3 didn't. Gemini 3.1 Pro (preview): Graded 4.0 of 5 on average (lowest 3). Kimi K3: Graded 4.3 of 5 on average (lowest 4); but doesn't end with a question.

  • Argue both sides of free buses (writing): Kimi K3 passed and Gemini 3.1 Pro (preview) didn't. Gemini 3.1 Pro (preview): Graded 3.7 of 5 on average (lowest 3). Kimi K3: Graded 4.7 of 5 on average (lowest 4).

  • An article in three bullets (summarization): Gemini 3.1 Pro (preview) passed and Kimi K3 didn't. Gemini 3.1 Pro (preview): Graded 4.3 of 5 on average (lowest 4). Kimi K3: Graded 4.0 of 5 on average (lowest 3); but 62 words, over the 60 allowed.

  • An email thread in one sentence (summarization): Gemini 3.1 Pro (preview) passed and Kimi K3 didn't. Gemini 3.1 Pro (preview): Graded 4.7 of 5 on average (lowest 4). Kimi K3: Graded 5.0 of 5 on average (lowest 5); but 32 words, over the 30 allowed.

  • A refund request outside the window (customer support): Gemini 3.1 Pro (preview) passed and Kimi K3 didn't. Gemini 3.1 Pro (preview): Graded 4.7 of 5 on average (lowest 4). Kimi K3: Graded 4.0 of 5 on average (lowest 2).

Job by job, the widest gaps first

  • Summarization: Gemini 3.1 Pro (preview) passed 5 of 5 and Kimi K3 3 of 5. Kimi K3 answered 2.4× sooner at the median, 8.0 s against 3.4 s. Kimi K3 cost 2.6× less, $0.0441 against $0.0172 for the 5 replies. Kimi K3's replies ran 36% longer, in tokens of reply, thinking not counted.

  • Customer support: Gemini 3.1 Pro (preview) passed 5 of 5 and Kimi K3 4 of 5. Kimi K3 answered 2.0× sooner at the median, 8.7 s against 4.3 s. Kimi K3 cost 2.3× less, $0.0529 against $0.0228 for the 5 replies. Kimi K3's replies ran 15% longer, in tokens of reply, thinking not counted.

  • RAG and answering from documents: Gemini 3.1 Pro (preview) passed 5 of 5 and Kimi K3 5 of 5. Kimi K3 answered 2.1× sooner at the median, 5.6 s against 2.6 s. Kimi K3 cost 4.0× less, $0.0290 against $0.0073 for the 5 replies. Kimi K3's replies ran 90% longer, in tokens of reply, thinking not counted.

  • Translation: Gemini 3.1 Pro (preview) passed 5 of 5 and Kimi K3 5 of 5. Kimi K3 answered 2.0× sooner at the median, 10.3 s against 5.2 s. Kimi K3 cost 3.9× less, $0.0699 against $0.0179 for the 5 replies. Kimi K3's replies ran 107% longer, in tokens of reply, thinking not counted.

  • SQL: Gemini 3.1 Pro (preview) passed 5 of 5 and Kimi K3 5 of 5. Kimi K3 answered 4.7× sooner at the median, 7.4 s against 1.6 s. Kimi K3 cost 3.3× less, $0.0386 against $0.0118 for the 5 replies. Their replies ran to about the same length.

  • Coding: Gemini 3.1 Pro (preview) passed 5 of 5 and Kimi K3 5 of 5. Kimi K3 answered 2.5× sooner at the median, 10.3 s against 4.1 s. Kimi K3 cost 3.1× less, $0.0973 against $0.0316 for the 5 replies. Their replies ran to about the same length.

  • Math: Gemini 3.1 Pro (preview) passed 5 of 5 and Kimi K3 5 of 5. Kimi K3 answered 1.5× sooner at the median, 8.3 s against 5.4 s. Kimi K3 cost 2.7× less, $0.0501 against $0.0187 for the 5 replies. Gemini 3.1 Pro (preview)'s replies ran 164% longer, in tokens of reply, thinking not counted.

  • Writing: Gemini 3.1 Pro (preview) passed 4 of 5 and Kimi K3 4 of 5. Kimi K3 answered 3.1× sooner at the median, 9.1 s against 3.0 s. Kimi K3 cost 2.4× less, $0.0459 against $0.0192 for the 5 replies. Kimi K3's replies ran 12% longer, in tokens of reply, thinking not counted.

  • Data analysis: Gemini 3.1 Pro (preview) passed 5 of 5 and Kimi K3 5 of 5. Kimi K3 answered 2.5× sooner at the median, 8.6 s against 3.5 s. Kimi K3 cost 2.3× less, $0.0773 against $0.0339 for the 5 replies. Gemini 3.1 Pro (preview)'s replies ran 94% longer, in tokens of reply, thinking not counted.

  • Agents and tool use: Gemini 3.1 Pro (preview) passed 5 of 5 and Kimi K3 5 of 5. Kimi K3 answered 2.7× sooner at the median, 5.3 s against 2.0 s. Kimi K3 cost 2.1× less, $0.0284 against $0.0135 for the 5 replies. Gemini 3.1 Pro (preview)'s replies ran 41% longer, in tokens of reply, thinking not counted.

All 50 prompts: who passed, who answered sooner, who cost less
Gemini 3.1 Pro (preview) and Kimi K3 on each prompt of our test runs
PromptResultSoonerCheaper
Turn a title into a URL slugBoth passedKimi K3, 2.4×took 6.0 s and 2.6 sKimi K3, 1.6×cost $0.0056 and $0.0035
Parse a duration like “1h 30m”Both passedKimi K3, 1.4×took 10.3 s and 7.1 sKimi K3, 2.7×cost $0.0172 and $0.0063
Merge overlapping intervalsBoth passedKimi K3, 4.2×took 8.6 s and 2.0 sKimi K3, 2.9×cost $0.0103 and $0.0036
Evaluate an arithmetic expression, no evalBoth passedKimi K3, 2.1×took 20.6 s and 9.9 sKimi K3, 3.1×cost $0.0351 and $0.0113
Parse CSV with quoted fieldsBoth passedKimi K3, 5.0×took 20.6 s and 4.1 sKimi K3, 4.2×cost $0.0291 and $0.0069
Announce a second bakery shop on LinkedInBoth passedKimi K3, 2.4×took 9.5 s and 3.9 sKimi K3, 2.5×cost $0.0113 and $0.0046
Rewrite corporate jargon in plain wordsBoth passedKimi K3, 4.4×took 8.8 s and 2.0 sKimi K3, 3.3×cost $0.0101 and $0.0030
Decline a meeting and offer two timesBoth passedKimi K3, 2.1×took 6.2 s and 3.0 sKimi K3, 1.7×cost $0.0058 and $0.0034
A product announcement with five rulesOnly Gemini 3.1 Pro (preview)Kimi K3, 4.0×took 9.2 s and 2.3 sKimi K3, 2.9×cost $0.0095 and $0.0033
Argue both sides of free busesOnly Kimi K3Kimi K3, 2.6×took 9.1 s and 3.6 sKimi K3, 1.9×cost $0.0092 and $0.0049
A discount, then sales taxBoth passedGemini 3.1 Pro (preview), 1.4×took 3.6 s and 5.0 sKimi K3, 1.1×cost $0.0042 and $0.0037
Pens at 3 for $4Both passedKimi K3, 1.8×took 9.6 s and 5.4 sKimi K3, 2.6×cost $0.0121 and $0.0046
Compound interest over three yearsBoth passedClosetook 6.2 s and 5.9 sKimi K3, 2.2×cost $0.0099 and $0.0044
Four-digit numbers whose digits sum to 9Both passedKimi K3, 1.6×took 9.6 s and 6.0 sKimi K3, 4.6×cost $0.0150 and $0.0032
The highest of three dice is a 5Both passedKimi K3, 3.4×took 8.3 s and 2.4 sKimi K3, 3.3×cost $0.0090 and $0.0027
An article in three bulletsOnly Gemini 3.1 Pro (preview)Kimi K3, 2.8×took 8.0 s and 2.8 sKimi K3, 3.1×cost $0.0088 and $0.0028
An email thread in one sentenceOnly Gemini 3.1 Pro (preview)Closetook 6.3 s and 6.4 sKimi K3, 2.0×cost $0.0058 and $0.0030
Decisions and action items from a meetingBoth passedKimi K3, 2.3×took 7.6 s and 3.4 sKimi K3, 1.9×cost $0.0081 and $0.0044
A quarterly memo for the CEOBoth passedGemini 3.1 Pro (preview), 1.4×took 11.2 s and 15.7 sKimi K3, 2.8×cost $0.0139 and $0.0049
A study with a negative resultBoth passedKimi K3, 2.5×took 8.2 s and 3.2 sKimi K3, 3.5×cost $0.0075 and $0.0021
The region with the most revenueBoth passedKimi K3, 3.8×took 8.5 s and 2.2 sKimi K3, 2.4×cost $0.0117 and $0.0049
Average order value in AugustBoth passedKimi K3, 3.2×took 8.6 s and 2.7 sKimi K3, 2.4×cost $0.0130 and $0.0054
Revenue change from July to AugustBoth passedKimi K3, 1.4×took 9.0 s and 6.4 sKimi K3, 2.1×cost $0.0135 and $0.0064
A median, filtered two waysBoth passedKimi K3, 2.0×took 7.1 s and 3.5 sKimi K3, 2.0×cost $0.0082 and $0.0042
Correlation between ad spend and sign-upsBoth passedGemini 3.1 Pro (preview), 1.2×took 16.5 s and 20.2 sKimi K3, 2.4×cost $0.0310 and $0.0131
A late orderBoth passedKimi K3, 3.5×took 8.7 s and 2.5 sKimi K3, 2.7×cost $0.0125 and $0.0045
A return inside the windowBoth passedKimi K3, 1.6×took 6.9 s and 4.3 sKimi K3, 2.3×cost $0.0076 and $0.0033
A frustrated customerBoth passedKimi K3, 1.7×took 9.8 s and 5.6 sKimi K3, 1.7×cost $0.0110 and $0.0063
A refund request outside the windowOnly Gemini 3.1 Pro (preview)Kimi K3, 2.8×took 7.4 s and 2.6 sKimi K3, 2.3×cost $0.0103 and $0.0045
A message with a planted instructionBoth passedClosetook 10.6 s and 11.5 sKimi K3, 2.8×cost $0.0116 and $0.0041
A delivery message into SpanishBoth passedKimi K3, 4.0×took 11.8 s and 2.9 sKimi K3, 5.0×cost $0.0157 and $0.0031
A product description into FrenchBoth passedKimi K3, 1.7×took 9.1 s and 5.4 sKimi K3, 3.9×cost $0.0141 and $0.0037
A meeting note into GermanBoth passedKimi K3, 4.4×took 13.0 s and 3.0 sKimi K3, 11.4×cost $0.0158 and $0.0014
Idioms into natural JapaneseBoth passedKimi K3, 2.0×took 10.3 s and 5.2 sKimi K3, 2.2×cost $0.0118 and $0.0053
A lease clause into Brazilian PortugueseBoth passedGemini 3.1 Pro (preview), 1.3×took 8.1 s and 10.1 sKimi K3, 2.8×cost $0.0125 and $0.0044
Customers in one countryBoth passedKimi K3, 4.5×took 4.9 s and 1.1 sKimi K3, 1.7×cost $0.0029 and $0.0018
Count orders by statusBoth passedKimi K3, 3.3×took 5.0 s and 1.5 sKimi K3, 4.5×cost $0.0039 and $0.0009
Revenue by categoryBoth passedKimi K3, 4.7×took 7.4 s and 1.6 sKimi K3, 3.9×cost $0.0074 and $0.0019
Every customer, even those without ordersBoth passedKimi K3, 3.1×took 7.9 s and 2.5 sKimi K3, 2.9×cost $0.0089 and $0.0031
Monthly revenue with a running totalBoth passedKimi K3, 4.4×took 11.1 s and 2.5 sKimi K3, 3.7×cost $0.0155 and $0.0042
A fact from one sectionBoth passedGemini 3.1 Pro (preview), 1.2×took 5.3 s and 6.5 sKimi K3, 4.5×cost $0.0045 and $0.0010
Core hours and start timesBoth passedKimi K3, 4.8×took 5.8 s and 1.2 sKimi K3, 4.7×cost $0.0052 and $0.0011
Two sections in one answerBoth passedKimi K3, 2.1×took 5.6 s and 2.6 sKimi K3, 2.4×cost $0.0047 and $0.0020
A later amendment changes the answerBoth passedKimi K3, 1.8×took 8.6 s and 4.8 sKimi K3, 4.5×cost $0.0099 and $0.0022
A question the handbook doesn't answerBoth passedKimi K3, 6.0×took 5.6 s and 0.9 sKimi K3, 4.6×cost $0.0046 and $0.0010
Pick the tool and work out the dateBoth passedKimi K3, 1.1×took 4.3 s and 3.7 sKimi K3, 1.5×cost $0.0048 and $0.0032
Convert a currencyBoth passedKimi K3, 1.8×took 5.3 s and 2.9 sKimi K3, 1.7×cost $0.0044 and $0.0025
Book a meeting from a sentenceBoth passedKimi K3, 4.5×took 7.1 s and 1.6 sKimi K3, 2.2×cost $0.0068 and $0.0031
Search, but don't bookBoth passedKimi K3, 2.4×took 4.7 s and 2.0 sKimi K3, 1.8×cost $0.0061 and $0.0034
Two calls with a unit conversionBoth passedKimi K3, 4.0×took 6.9 s and 1.7 sKimi K3, 5.1×cost $0.0063 and $0.0012

Gemini 3.1 Pro (preview) vs Kimi K3 in our test runs

Gemini 3.1 Pro (preview) and Kimi K3 on the same prompts, job by job: how many replies passed their check.

Based on 100 of our test runs on , through OpenRouter with the app's own prompt and settings. Updated .

Gemini 3.1 Pro (preview) and Kimi K3 in our test runs, job by job
JobGemini 3.1 Pro (preview)Kimi K3
Coding5 of 55 of 5
Writing4 of 54 of 5
Math5 of 55 of 5
Summarization5 of 53 of 5
Data analysis5 of 55 of 5
Customer support5 of 54 of 5
Translation5 of 55 of 5
SQL5 of 55 of 5
RAG and answering from documents5 of 55 of 5
Agents and tool use5 of 55 of 5
All jobs49 of 50 · 1 message each · $0.0107 a reply46 of 50 · 1 message each · $0.0039 a reply

The same prompts, sent to both through llmwise's own pipeline and checked the same way: five for each job. A bold count is the one that passed more. Messages: what each reply counted as on Pro. Cost: what OpenRouter charged us per reply, on average. Every prompt, and how it's scored.

Where they differed most, summarization: both replies to a prompt only one of them passed.

An everyday prompt: “An article in three bullets”, in full.

  • Gemini 3.1 Pro

    Passed: Graded 4.3 of 5 on average (lowest 4).

    * Daily bike trips jumped to 2,300 and cyclist injuries dropped sharply.
    * Car traffic fell 14% with minor delays, while local business impacts were mixed.
    * The city council votes in November on spending €1.1 million to make the lanes permanent.

    560 tokens in, 636 out (577 of them reasoning) · 8.0 s · $0.0088 · 1 message on Pro · answered by google/gemini-3.1-pro-preview via Google ·

  • Kimi K3

    Failed: Graded 4.0 of 5 on average (lowest 3); but 62 words, over the 60 allowed.

    - Aldmoor's six-month protected bike lane trial on Market Street ended; the council votes in November on making them permanent.
    - Bike trips rose from 900 to 2,300 daily, cyclist injuries dropped from nine to two, while car traffic fell 14% with only slightly longer rush-hour drive times.
    - Businesses were split on customer impact; installation cost €410,000, with permanent lanes requiring €1.1 million more.

    608 tokens in, 109 out (10 of them reasoning) · 2.8 s · $0.0028 · 1 message on Pro · answered by moonshotai/kimi-k3 via Phala ·

Gemini 3.1 Pro (preview) and Kimi K3 on every plan

Whether the one-time free trial reaches each model, then each paid plan's messages on it.

Messages on Gemini 3.1 Pro (preview) and Kimi K3, plan by plan
PlanPriceGemini 3.1 Pro (preview)Kimi K3
Free$0In the one-time trial of 5 messagesIn the one-time trial of 5 messages
Pro$20 a monthUp to 125 a monthUp to 125 a month
Max$50 a monthUp to 400 a monthUp to 400 a month
Ultra$100 a monthUp to 900 a monthUp to 900 a month
Studio$200 a monthUp to 2,000 a monthUp to 2,000 a month

Prices don't include tax, which is added where it applies and shown before you pay. A paid plan's month is one allowance shared by every model, so each monthly count is the most you get if all of it goes to that model. It renews each billing period; everyday models refill daily at 00:00 UTC. Long chats count more per reply. How pricing works.

Every limit is published. Paid plans also have a monthly fair-use limit on AI cost: Pro $7.50, Max $20, Ultra $42, Studio $85. Using every message on your plan at typical sizes stays under it; very large messages and heavy research use it faster. Every limit, explained.

What differs

  • Messages on Pro

    They cost the same in llmwise: up to 125 messages a month on Pro on either.

  • Context window

    Both take up to 1.05M tokens of context. A chat in llmwise holds up to 200k tokens, which fits in either, so the difference shows only through each maker's own API.

  • Images and PDFs

    Both read images. Kimi K3 gets a PDF's text rather than the file itself.

  • On Free

    Both are in the free trial.

  • Where messages go

    Gemini 3.1 Pro (preview): Sent to Google directly. Kimi K3: Served through OpenRouter, only by hosts that don't store or train on prompts. The maker's own endpoint is never asked.

Fact by fact

Gemini 3.1 Pro (preview) and Kimi K3, fact by fact
FactGemini 3.1 Pro (preview)Kimi K3
Context window1.05M tokens1.05M tokens
Reads imagesYesYes
PDFsWhole fileText only
ReasoningYesYes
API price (September 2026)$2.00 in / $12.00 out per million tokens$3.00 in / $15.00 out per million tokens
A typical message at API prices (4,000 tokens in, 700 out)$0.0164$0.0225
A $10 top-up adds100 messages100 messages
Where a message goesSent to Google directly.Served through OpenRouter, only by hosts that don't store or train on prompts. The maker's own endpoint is never asked.
If the provider failsIf Google fails before the reply starts (an overload, a server error, a dropped connection), llmwise sends the same request to Gemini 3.1 Pro (preview) through OpenRouter instead.When one host is down, OpenRouter moves the request to another host that meets the same rules.
Anthropic's safety fallbackDoesn't applyDoesn't apply
API prices are what our model catalog lists (Gemini: Google's list price; Kimi: Moonshot's list price). In llmwise you pay per message, not per token: the counts above are what you get.

Gemini 3.1 Pro (preview) or Kimi K3?

From the facts above and our test runs: the rest is how their answers suit your work, which one chat can show you.

  • Pick Gemini 3.1 Pro (preview): it reads a PDF as the whole file, charts and scans included; it costs its maker less to run ($0.0164 a typical message at API prices), though in llmwise the count is the same.

Each model's page, the families, and other pairs

Gemini 3.1 Pro (preview) vs Kimi K3 is one pair of models. The page below covers the whole families.

Questions

Is Gemini 3.1 Pro (preview) or Kimi K3 cheaper in llmwise?

They cost the same in llmwise: up to 125 messages a month on Pro on either. Every paid plan's monthly allowance is shared by all models, so each count is the most you get if it all goes to that model.

Can I try Gemini 3.1 Pro (preview) and Kimi K3 for free?

Yes: both are in the free trial of 5 messages.

Which has the bigger context window, Gemini 3.1 Pro (preview) or Kimi K3?

Neither: both take 1.05M tokens. A chat in llmwise holds up to 200k tokens, which fits in either, so the difference shows only through each maker's own API.

Can I use Gemini 3.1 Pro (preview) and Kimi K3 in the same chat?

Yes. Pick Gemini 3.1 Pro (preview) for one message and Kimi K3 for the next; the second sees the whole chat, including the first one's answer.

Which did better in your test runs, Gemini 3.1 Pro (preview) or Kimi K3?

On the same 50 prompts, run on September 27, 2026, Gemini 3.1 Pro (preview) passed 49 and Kimi K3 passed 46. The table on this page has each job, and every reply is published.

Claude, GPT, Gemini, DeepSeek, Grok, Kimi, and GLM, in one chat.

See what a message costs before you send it. Free is 5 messages to try; sign in with an email link, no password or card.