BroadRouter

Kimi K3 API pricing

Updated

Among the 18 providers we checked (prices read 2026-11-14 to 2026-11-15), the lowest blended standard price for Kimi K3 is Chutes at USD 1.85 per 1M input tokens and USD 9.50 per 1M output tokens. The median across 18 standard prices is USD 3 / USD 15. Moonshot AI's own API charges USD 3 / USD 15.

Moonshot AI · slug kimi-k3 · up to 1,048,576 tokens of context · 18 providers with a standard price. Every price is read from the provider's own pricing page and linked in its row, checked 2026-11-15. No aggregator feed. Download all prices (every provider and model, with sources, CSV or JSON). How it compares with the market: the LLM API price index. Archived list prices since 2026-08-10 (7 archived price points, 0 list-price changes): LLM API price history.

What does Moonshot AI charge for Kimi K3?

Input
USD 3
Cached input
USD 0.30
Output
USD 15
Blended 3:1
USD 6

USD per 1M tokens on Moonshot AI (Kimi), model kimi-k3, standard tier. Source, checked 2026-11-14.

Other Moonshot AI tiers: batch USD 1.80 / USD 9.

Is a host cheaper than Moonshot AI? Yes. Chutes charges USD 1.85 input and USD 9.50 output per 1M tokens, 37% below Moonshot AI's own USD 3 / USD 15 on the 3:1 blend.

Kimi K3 price per provider

USD per 1M tokens. Standard prices are ranked on a blend of 3 input tokens to 1 output token, cheapest first; the dearest standard price is 99% above the cheapest. Free rows, the maker's serving version and other service tiers follow, tagged and unranked.

Kimi K3 API prices by provider
#ProviderInputCached inputOutputBlended 3:1ContextPrecisionSource
1Chutes cheapest
moonshotai/Kimi-K3-TEE
Notes and source quote

Runs on confidential TEE compute.

“Kimi-K3 moonshotai/Kimi-K3-TEE $0.185 per 1M tokens $1.85 per 1M tokens $9.50 per 1M tokens 1.0M”
USD 1.85USD 0.185USD 9.50USD 3.7631,048,576MXFP4Chutes page
checked 2026-11-15
re-checked 2026-11-16
2Featherless
moonshotai/Kimi-K3
Notes and source quote

Developer plan (per-token billing, from 50 USD a month in credits). The page marks this price with "‡" (a pricing addendum shown only on hover); Featherless's public model API lists 3 input / 15 output for the same model, so this price may be temporary.

“PRICING: INPUT $2 / CACHED $0.3 / OUTPUT $10”
USD 2USD 0.30USD 10USD 4262,144MXFP4Featherless page
checked 2026-11-15
re-checked 2026-11-16
3SiliconFlow
Notes and source quote

Context shown as 1049K.

“Kimi-K3 1049K $2.7 $0.27 $13.5”
USD 2.70USD 0.27USD 13.50USD 5.401,048,576not statedSiliconFlow page
checked 2026-11-14
re-checked 2026-11-16
4DeepInfra
moonshotai/Kimi-K3
Notes and source quote
“$2.85 in $14.25 out $0.285 cached / 1M tokens”
USD 2.85USD 0.285USD 14.25USD 5.701,048,576MXFP4DeepInfra page
checked 2026-11-14
re-checked 2026-11-16
5Moonshot AI maker
kimi-k3
Notes and source quote

Cache writes are billed separately (5m and 1h TTL).

“kimi-k3 | $3.00 / MTok | $15.00 / MTok | $0.30 / MTok | $3.00 / MTok | $6.00 / MTok”
USD 3USD 0.30USD 15USD 61,048,576not statedMoonshot AI page
checked 2026-11-14
re-checked 2026-11-16
6Alibaba Cloud
kimi-k3
Notes and source quote

International (Singapore) deployment. Singapore tab of the Kimi table.

“kimi-k3 International $3 $15”
USD 3not publishedUSD 15USD 6not statednot statedAlibaba Cloud page
checked 2026-11-14
re-checked 2026-11-16
7Baseten
moonshotai/Kimi-K3
Notes and source quote

Context from the model library page (https://www.baseten.co/library/kimi-k3/), stated as 1M.

“Kimi K3 $3.00 $0.30 $15.00”
USD 3USD 0.30USD 15USD 61,000,000not statedBaseten page
checked 2026-11-14
re-checked 2026-11-16
8DigitalOcean
Notes and source quote

DigitalOcean-hosted.

“Kimi K3 — Processing modePrompt lengthInputOutputCache read: Standard All prompts$3.00$15.00$0.30”
USD 3USD 0.30USD 15USD 6not statednot statedDigitalOcean page
checked 2026-11-14
re-checked 2026-11-16
9Fireworks AI
accounts/fireworks/models/kimi-k3
Notes and source quote
“Kimi K3 | $3.00 / $0.30 / $15.00 | $3.75 / $0.375 / $18.75”
USD 3USD 0.30USD 15USD 61,048,576not statedFireworks AI page
checked 2026-11-14
re-checked 2026-11-16
10GMI Cloud
moonshotai/kimi-k3
Notes and source quote
“Price (per 1M tokens) Input $3.00 Output $15.00 Cache Read $0.30 Cache Write -”
USD 3USD 0.30USD 15USD 61,048,576FP4GMI Cloud page
checked 2026-11-15
re-checked 2026-11-16
11Hyperbolic
kimi-k3
Notes and source quote

Context shown on page as "1M" (rounded); stored as the decimal value.

“Kimi K3 kimi-k3 Moonshot AI · 1M context · Text $3.00 $15.00 $0.30”
USD 3USD 0.30USD 15USD 61,000,000not statedHyperbolic page
checked 2026-11-14
re-checked 2026-11-16
12Microsoft Azure
Notes and source quote

Foundry Models serverless, served by Fireworks on Azure, Global deployment.

“FW Kimi K3 Global Input: $3/1M Tokens | Cached Input: $0.30/1M Tokens | Output: $15/1M Tokens”
USD 3USD 0.30USD 15USD 6not statednot statedMicrosoft Azure page
checked 2026-11-14
re-checked 2026-11-16
13Novita AI
moonshotai/kimi-k3
Notes and source quote

Cache write (5m) 3 per 1M.

“Kimi K3 1M $3 /Mt · Cache Read $0.3 /Mt · Cache Write(5m) $3 /Mt · Cache Write(1h) $6 /Mt $15 /Mt”
USD 3USD 0.30USD 15USD 61,048,576not statedNovita AI page
checked 2026-11-14
re-checked 2026-11-16
14Parasail
Notes and source quote
“Kimi K3 $3.00 $15.00 $0.30”
USD 3USD 0.30USD 15USD 6not statednot statedParasail page
checked 2026-11-14
re-checked 2026-11-16
15Perplexity
perplexity/kimi-k3
Notes and source quote

Hosted by Perplexity ("Perplexity-Hosted Models"); tool calls billed separately.

“`perplexity/kimi-k3` | 3.00 | 15.00 | 0.30”
USD 3USD 0.30USD 15USD 6not statednot statedPerplexity page
checked 2026-11-15
re-checked 2026-11-16
16Together AI
Notes and source quote

US geo-pinned option: $3.75 in / $18.75 out.

“Kimi K3 $3.00 $0.30 (cached) $15.00”
USD 3USD 0.30USD 15USD 6not statednot statedTogether AI page
checked 2026-11-14
re-checked 2026-11-16
17Amazon Bedrock
Notes and source quote

US cross-region (US CRIS) row; Global CRIS is $3.00 / $15.00 / cache read $0.30.

“Kimi 3 … US CRIS $3.30 $16.50 $0.33 $4.125”
USD 3.30USD 0.33USD 16.50USD 6.60not statednot statedAmazon Bedrock page
checked 2026-11-14
re-checked 2026-11-16
18Venice
kimi-k3
Notes and source quote
“| Kimi K3 | `kimi-k3` | $3.75 | $18.75 | $0.375 | — | Private |”
USD 3.75USD 0.375USD 18.75USD 7.501,000,000not statedVenice page
checked 2026-11-15
re-checked 2026-11-16
—Moonshot AI maker batch
kimi-k3
Notes and source quote

Batch API, billed at 60% of the standard price. Cache writes are billed separately (5m and 1h TTL).

“kimi-k3 | $1.80 / MTok | $9.00 / MTok | $0.18 / MTok | $1.80 / MTok | $3.60 / MTok”
USD 1.80USD 0.18USD 9USD 3.601,048,576not statedMoonshot AI page
checked 2026-11-14
re-checked 2026-11-16
—Fireworks AI flex / spot
accounts/fireworks/models/kimi-k3
Notes and source quote

Fireworks "Spot" mode: 30% off Standard on best-effort capacity; throughput can drop to zero.

“Kimi K3 Spot $0.21 $2.10 $10.50”
USD 2.10USD 0.21USD 10.50USD 4.201,048,576not statedFireworks AI page
checked 2026-11-14
re-checked 2026-11-16
—Fireworks AI priority
accounts/fireworks/models/kimi-k3
Notes and source quote

Priority mode.

“Kimi K3 | Priority $3.75 / $0.375 / $18.75”
USD 3.75USD 0.375USD 18.75USD 7.501,048,576not statedFireworks AI page
checked 2026-11-14
re-checked 2026-11-16
—Fireworks AI fast
accounts/fireworks/models/kimi-k3
Notes and source quote

Fast mode.

“Kimi K3 Fast | $4.50 / $0.45 / $22.50 | —”
USD 4.50USD 0.45USD 22.50USD 91,048,576not statedFireworks AI page
checked 2026-11-14
re-checked 2026-11-16
—Venice fast
kimi-k3-fast-api
Notes and source quote

Fast variant, model ID kimi-k3-fast-api.

“| Kimi K3 Fast | `kimi-k3-fast-api` | $4.50 | $22.50 | $0.45 | — | Private |”
USD 4.50USD 0.45USD 22.50USD 91,000,000not statedVenice page
checked 2026-11-15
re-checked 2026-11-16

Report a wrong price or number (email to info@broadnet.ai, with this page's link filled in).

What does Kimi K3 cost for a real workload?

Monthly cost over 30 days at each standard price, using the presets of the LLM API cost calculator. Providers without a cached price bill cached input at the input price.

  • Support chatbot: 2,000 requests a day, 1,500 input tokens (60% cached), 300 output.
  • RAG over documents: 1,000 requests a day, 8,000 input tokens (30% cached), 500 output.
  • Coding agent (one developer): 1,500 requests a day, 40,000 input tokens (85% cached), 2,000 output incl. reasoning.
Monthly cost of three workloads on Kimi K3, by provider
ProviderSupport chatbotRAG over documentsCoding agent (one developer)
ChutesUSD 247.59USD 466.62USD 1,638
FeatherlessUSD 268.20USD 507.60USD 1,899
SiliconFlowUSD 354.78USD 675.54USD 2,357
DeepInfraUSD 374.49USD 713.07USD 2,488
Moonshot AI (maker)USD 394.20USD 750.60USD 2,619
Alibaba CloudUSD 540.00USD 945.00USD 6,750
BasetenUSD 394.20USD 750.60USD 2,619
DigitalOceanUSD 394.20USD 750.60USD 2,619
Fireworks AIUSD 394.20USD 750.60USD 2,619
GMI CloudUSD 394.20USD 750.60USD 2,619
HyperbolicUSD 394.20USD 750.60USD 2,619
Microsoft AzureUSD 394.20USD 750.60USD 2,619
Novita AIUSD 394.20USD 750.60USD 2,619
ParasailUSD 394.20USD 750.60USD 2,619
PerplexityUSD 394.20USD 750.60USD 2,619
Together AIUSD 394.20USD 750.60USD 2,619
Amazon BedrockUSD 433.62USD 825.66USD 2,881
VeniceUSD 492.75USD 938.25USD 3,274

For the support chatbot, the cheapest price (Chutes, USD 247.59 a month) saves USD 146.61 against the median (USD 394.20). For the coding agent the gap is USD 981.45 a month (37% lower).

Which Kimi K3 provider is cheapest for your traffic?

Chutes has the lowest input price (USD 1.85) and the lowest output price (USD 9.50), so it is the cheapest standard price at every mix of input and output tokens, before cache discounts.

Kimi K3: price per 1M tokens at 3 input to 1 output, by share of input served from cache
ProviderNo cache hits50% cached90% cached
ChutesUSD 3.763 cheapestUSD 3.138 cheapestUSD 2.639 cheapest
FeatherlessUSD 4USD 3.362USD 2.853
SiliconFlowUSD 5.40USD 4.489USD 3.76
DeepInfraUSD 5.70USD 4.738USD 3.969
Moonshot AI (maker)USD 6USD 4.987USD 4.178
Alibaba Cloud (no cache price)USD 6USD 6USD 6

USD per 1M tokens, 3 input tokens to 1 output token, each provider's cheapest standard row; 12 dearer providers are in the price table above. A provider without a cached price bills cached input at the input price. Any charge for writing to the cache, which some providers add, is not included.

17 of 18 providers publish a cached-input price below the input price: Chutes (90% off), Featherless (85% off), SiliconFlow (90% off), DeepInfra (90% off), Moonshot AI (90% off) and 12 more. Alibaba Cloud publishes none, so cached tokens there bill at the input price.

Context windows differ: 1,048,576 tokens at Chutes, SiliconFlow, DeepInfra and 4 more; 1,000,000 tokens at Baseten, Hyperbolic and Venice; 262,144 tokens at Featherless. Alibaba Cloud, DigitalOcean, Microsoft Azure, Parasail, Perplexity and 2 more do not state one. The cheapest row is not always the one that fits a long prompt.

What does Kimi K3 cost in other languages?

The same text needs a different number of tokens in each language. With the Kimi K3 tokenizer, sending 1M characters costs this much at the cheapest standard input price we found for Kimi K3:

ContentInput per 1M charactersCheapest provider
EnglishUSD 0.402Chutes
Arabic MSAUSD 0.887Chutes
HindiUSD 1.05Chutes
ChineseUSD 1.01Chutes
CodeUSD 0.480Chutes

From BroadRouter's real cost by language dataset, which covers 15 languages and content types for every model we could match to a measured tokenizer.

Published benchmark results

BroadNet did not run these benchmarks; results are as published by the Moonshot AI model card and Epoch AI and may use different settings. BroadRouter is BroadNet's site.

Published benchmark results for Kimi K3, with sources
BenchmarkResultSetting or noteSource
GPQA Diamond93.5Max reasoning settingMoonshot AI model card, checked 2026-11-16
Terminal-Bench 2.188.3Max reasoning settingMoonshot AI model card, checked 2026-11-16
DeepSWE67.5Max reasoning settingMoonshot AI model card, checked 2026-11-16
BrowseComp91.2Max reasoning settingMoonshot AI model card, checked 2026-11-16
GPQA Diamond0.931Max reasoning setting; Epoch run kimi-k3_max, mean score on a 0 to 1 scaleEpoch AI, checked 2026-11-16

Epoch AI rows: Epoch AI, Capabilities & benchmarking, licensed CC BY 4.0, shown as the mean score on a 0 to 1 scale, rounded to three decimals (hover a value for the exact figure in Epoch's data file).

Estimate your own bill

The Moonshot AI models whose cheapest standard price is closest to Kimi K3's (USD 3.763 blended):

Moonshot AI models priced closest to Kimi K3
ModelCheapest inputCheapest outputMax contextProviders
Kimi K3.1 ProUSD 1.90USD 9.501,048,57613
Kimi K2.7 CodeUSD 0.60USD 3262,14412
Kimi K2 Instruct 0905USD 0.60USD 2.50262,1442
Kimi K2 InstructUSD 0.57USD 2.30131,0722
Kimi K2 ThinkingUSD 0.80USD 1.20262,1446
Kimi K2.5USD 0.45USD 2.25262,1448

Priced closest from other makers: Claude Sonnet 5 (Anthropic, USD 4) · Claude Sonnet 5.5 (Anthropic, USD 4) · Gemini 4 Argon (Google, USD 4).

Compare Kimi K3 with

Also compared: vs Kimi K3.1 Pro · vs Kimi K2.6 · vs Claude Sonnet 5.5 · vs Qwen3.8 Max · vs DeepSeek V4 Pro · vs GLM 5.3 · vs MiniMax M3.

Every Kimi K3 price comparison · All model-vs-model price comparisons

All 428 models: API prices compared · Cheapest LLM APIs right now · LLM API price index (open data) · Download all prices (CSV, JSON) · Price change log

Questions

How much does the Kimi K3 API cost per token?

From USD 1.85 per 1M input tokens (USD 0.00000185 per token) and USD 9.50 per 1M output tokens (USD 0.0000095 per token) at the lowest standard prices we found. The median of 18 standard prices is USD 3 input and USD 15 output. Each price was read from the provider's own pricing page; the source is linked in its row.

Which provider is cheapest for Kimi K3?

Among the 18 providers we checked (prices read 2026-11-14 to 2026-11-15), the lowest is Chutes, ranked on a blend of 3 input tokens to 1 output token (USD 3.763 per 1M). It also has the lowest input and the lowest output price. Batch, flex, priority and fast tiers are shown in the table but not ranked.

Is Kimi K3 cheaper from a host than from Moonshot AI?

Yes. Chutes charges USD 1.85 input and USD 9.50 output per 1M tokens, 37% below Moonshot AI's own USD 3 / USD 15 on the 3:1 blend.

Does Kimi K3 have cached input pricing?

17 of 18 standard prices include one. The lowest is USD 0.185 per 1M at Chutes; the median is USD 0.30. Providers without one bill cached input at the input price. What cached input means.

What is the Kimi K3 context window?

Up to 1,048,576 tokens. 4 standard prices come with less, down to 262,144 tokens (Featherless, Baseten, Hyperbolic and Venice). Check the context column before you pick the cheapest row for long prompts. Context window calculator.

How much does a coding agent cost per month on Kimi K3?

For 1,500 requests a day, 40,000 input tokens (85% cached), 2,000 output incl. reasoning, over 30 days: USD 1,638 a month at Chutes, USD 2,619 at the median price and USD 6,750 at Alibaba Cloud. Try your own numbers.

Is there a batch, flex or priority price for Kimi K3?

Yes: batch from Moonshot AI at USD 1.80 / USD 9; flex / spot from Fireworks AI at USD 2.10 / USD 10.50; priority from Fireworks AI at USD 3.75 / USD 18.75; fast from Fireworks AI at USD 4.50 / USD 22.50 per 1M input / output tokens. These tiers trade speed or availability for price (batch runs within hours, flex and spot can be slow or refused at capacity, priority and fast cost more for speed or reliability), so they are shown but not ranked.

Is Kimi K3 free to use through an API?

No provider in our data lists a free price for Kimi K3 as of 2026-11-15. Free tiers and credits by provider (checked 2026-11-14) lists general offers that do not name a model; all free LLM API models are on one page.

How does Kimi K3 score on benchmarks?

As published, the Moonshot AI model card reports 93.5 on GPQA Diamond and 88.3 on Terminal-Bench 2.1; Epoch AI measured 0.931 on GPQA Diamond in its own run, as a mean score on a 0 to 1 scale (checked 2026-11-16). BroadNet did not run these benchmarks, and sources may use different settings, so the figures are not directly comparable. Every published result we found, with its setting and source.

Read as MarkdownAsk ChatGPTAsk ClaudeAsk Perplexity