Signal SentrySearchUpdated Nov 20, 05:30 UTCAM Drop
Models · scorecard

Kimi K3

What the vendor claims, next to what independent boards measured. The two are never mixed in one column.

At a glance

Maker
MoonshotAI
Price per 1M tokens
$0.192 in · $16 out
Context window
1.05M
Input → output
text, image, video → text
Added to OpenRouter
Jul 16, 2026

Source: OpenRouter, as of Nov 20, 06:00 UTC. Model page on OpenRouter

What the vendor claims

Quoted word for word from the vendor's own launch post or model card. Claims, not measurements.

No vendor claims recorded yet. They come from launch posts and model cards the Drops covered.

Independent results

Measured by independent sources (Epoch AI, LMArena, OpenRouter's benchmark runs, τ²-bench, Vectara), not by the vendor. "Listed as" is the source's own name for the model, so you can check our match.

Independent benchmark results for this model, each with its source and date
BoardResultRank
Epoch AI · MirrorCodeListed as: kimi-k3_maxSource: Epoch AI (CC BY), as of Nov 20, 2026. Source54.4%Accuracy18
Epoch AI · Furniture AssemblyListed as: kimi-k3_maxSource: Epoch AI (CC BY), as of Nov 20, 2026. Source34.2%Accuracy27
Epoch AI · Epoch Capabilities IndexListed as: Kimi K3Source: Epoch AI (CC BY), as of Nov 20, 2026. Source158.6Index25
LMArena · Text (style control)Listed as: Kimi K3 (Max)LMArena (CC BY 4.0), as of Nov 18, 2026. Source1,487Arena score21
LMArena · WebDevListed as: Kimi K3 (Max)LMArena (CC BY 4.0), as of Nov 18, 2026. Source1,658Arena score23
OpenRouter · GPQA DiamondListed as: Kimi K3via Moonshot AI, its own providerSource: OpenRouter (openrouter.ai), licensed under CC BY 4.0, as of Nov 20, 2026. Source91.9%Accuracy21
OpenRouter · GPQA Diamond: cost per taskListed as: Kimi K3via Moonshot AI, its own providerLower is better on this board.Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0, as of Nov 20, 2026. Source$0.14Cost per task47
OpenRouter · GPQA Diamond: time per taskListed as: Kimi K3via Moonshot AI, its own providerLower is better on this board.Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0, as of Nov 20, 2026. Source280 sTime per task42
OpenRouter · τ²-bench airlineListed as: Kimi K3via Moonshot AI, its own providerSource: OpenRouter (openrouter.ai), licensed under CC BY 4.0, as of Nov 20, 2026. Source73.6%Accuracy56
OpenRouter · τ²-bench airline: cost per taskListed as: Kimi K3via Moonshot AI, its own providerLower is better on this board.Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0, as of Nov 20, 2026. Source$0.29Cost per task42
OpenRouter · τ²-bench airline: time per taskListed as: Kimi K3via Moonshot AI, its own providerLower is better on this board.Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0, as of Nov 20, 2026. Source407 sTime per task50

Price history

OpenRouter's list price per 1M tokens in each of our snapshots.

  • Output now $16
  • Input now $0.192

Changed 33 times across our 93 snapshots.

Show the numbers
This model's OpenRouter price in each of our snapshots
SnapshotInput $/1MOutput $/1M
Nov 20, 06:04 UTC$0.192$16
Nov 19, 16:04 UTC$0.192$16
Nov 19, 04:04 UTC$0.192$16
Nov 18, 18:04 UTC$0.192$16
Nov 18, 04:03 UTC$0.192$16
Nov 17, 16:01 UTC$0.192$16
Nov 17, 04:04 UTC$0.192$16
Nov 16, 19:54 UTC$0.192$16
Nov 16, 04:05 UTC$0.192$16
Nov 15, 16:04 UTC$0.192$16
Nov 15, 04:03 UTC$0.192$16
Nov 14, 16:04 UTC$0.192$16
Nov 14, 04:05 UTC$0.192$16
Nov 13, 16:05 UTC$0.192$16
Nov 13, 04:04 UTC$0.192$16
Nov 12, 16:05 UTC$0.192$16
Nov 12, 04:04 UTC$0.192$16
Nov 11, 16:01 UTC$0.192$16
Nov 11, 04:04 UTC$0.192$16
Nov 10, 16:04 UTC$0.192$16
Nov 10, 04:05 UTC$0.192$16
Nov 9, 16:04 UTC$0.192$16
Nov 9, 04:04 UTC$0.192$16
Nov 8, 16:04 UTC$0.192$16
Nov 8, 04:03 UTC$0.192$16
Nov 7, 16:04 UTC$0.192$16
Nov 7, 04:04 UTC$0.192$16
Nov 6, 16:04 UTC$0.192$16
Nov 6, 04:04 UTC$0.192$16
Nov 5, 16:05 UTC$0.192$16
Nov 5, 10:05 UTC$0.192$16
Nov 4, 16:02 UTC$0.192$16
Nov 4, 04:04 UTC$0.2$16
Nov 3, 16:04 UTC$0.192$15
Nov 3, 13:26 UTC$0.192$15
Nov 2, 16:01 UTC$0.192$15
Nov 2, 04:04 UTC$0.192$15
Nov 1, 16:03 UTC$0.192$15
Nov 1, 04:04 UTC$0.192$15
Oct 31, 16:04 UTC$0.192$15
Oct 31, 04:04 UTC$0.192$15
Oct 30, 16:03 UTC$0.192$15
Oct 30, 04:04 UTC$0.192$15
Oct 29, 16:03 UTC$0.239$10.8
Oct 29, 04:04 UTC$0.192$20
Oct 28, 16:04 UTC$0.174$20
Oct 28, 09:22 UTC$0.179$20
Oct 27, 16:03 UTC$0.179$20
Oct 27, 04:03 UTC$0.2$20
Oct 26, 16:04 UTC$0.189$20
Oct 26, 04:03 UTC$0.152$20
Oct 25, 16:04 UTC$0.239$10.8
Oct 25, 06:03 UTC$0.239$10.8
Oct 24, 15:04 UTC$0.239$10.8
Oct 24, 03:03 UTC$0.239$10.8
Oct 23, 15:03 UTC$0.239$11.936
Oct 23, 03:04 UTC$0.239$11.936
Oct 22, 15:03 UTC$0.239$11.936
Oct 22, 05:04 UTC$0.239$11.936
Oct 21, 15:03 UTC$0.239$11.936
Oct 21, 03:02 UTC$0.239$11.936
Oct 20, 15:04 UTC$0.239$11.936
Oct 20, 03:03 UTC$0.239$11.936
Oct 19, 15:04 UTC$0.239$11.936
Oct 19, 10:50 UTC$0.239$11.936
Oct 18, 15:03 UTC$0.304$11.936
Oct 18, 03:05 UTC$0.304$11.936
Oct 17, 15:03 UTC$0.304$11.936
Oct 17, 03:03 UTC$0.304$11.936
Oct 16, 15:02 UTC$0.304$11.936
Oct 16, 03:03 UTC$0.405$15.45
Oct 15, 17:02 UTC$0.45$16.5
Oct 15, 03:02 UTC$0.45$15
Oct 14, 15:03 UTC$0.7$15
Oct 14, 03:02 UTC$0.7$13.5
Oct 13, 15:03 UTC$0.73$10
Oct 13, 05:03 UTC$0.73$13.5
Oct 12, 15:03 UTC$0.73$10
Oct 12, 03:04 UTC$0.73$10
Oct 11, 15:02 UTC$0.8$10
Oct 11, 03:03 UTC$0.8$12
Oct 10, 15:01 UTC$0.5$13.5
Oct 10, 03:04 UTC$0.64$13.5
Oct 9, 15:05 UTC$0.5$12
Oct 9, 03:04 UTC$0.6$12
Oct 8, 15:03 UTC$0.62$12.3
Oct 8, 03:05 UTC$0.65$15
Oct 7, 15:03 UTC$0.62$15
Oct 7, 03:03 UTC$0.69$15
Oct 6, 15:03 UTC$0.83$14
Oct 6, 03:02 UTC$0.99$14
Oct 5, 15:02 UTC$0.67$14
Oct 5, 03:04 UTC$0.67$14

Source: OpenRouter, as of Nov 20, 06:04 UTC

ValueOur calculation

Capability vs price

This model in colour, against every other model on the Epoch Capabilities Index that OpenRouter prices.

Scatter chart of 29 models, capability against the price of a sample task. Best value at its price: Claude Haiku 5.5, GPT-6.1 Terra, Gemini 4 Carbon, GPT-6.2 Astra.155160165170175$0.001$0.01$0.10$1.00Price of the sample task (log scale)Epoch Capabilities IndexGPT-6.2 Astra (OpenAI): index 174.17, $0.16 per sample taskClaude Fable 5.2 (Anthropic): index 173.5, $0.16 per sample taskGemini 4 Carbon (Google): index 169.78, $0.04 per sample taskGPT-6 Astra (OpenAI): index 169.2, $0.20 per sample taskClaude Opus 5.5 (Anthropic): index 168.7, $0.08 per sample taskGPT-6.1 Sol (OpenAI): index 168.63, $0.04 per sample taskGPT-6.1 Terra (OpenAI): index 167.72, $0.02 per sample taskClaude Fable 5.1 (Anthropic): index 167.09, $0.20 per sample taskClaude Sonnet 5.5 (Anthropic): index 166.59, $0.04 per sample taskGemini 4 Argon (Google): index 166.57, $0.04 per sample taskClaude Opus 5 (Anthropic): index 164.74, $0.10 per sample taskClaude Fable 5 (Anthropic): index 164.21, $0.20 per sample taskGPT-6 Sol (OpenAI): index 163.82, $0.04 per sample taskGPT-5.6 Sol (OpenAI): index 163.05, $0.08 per sample taskGPT-5.5 Pro (OpenAI): index 162.77, $0.66 per sample taskMuse Wonder (Meta): index 161.74, $0.08 per sample taskKimi K3.1 Pro (MoonshotAI): index 161.36, $0.02 per sample taskGPT-5.6 Terra (OpenAI): index 160.78, $0.044 per sample taskGPT-5.4 Pro (OpenAI): index 160.32, $0.66 per sample taskClaude Haiku 5.5 (Anthropic): index 160.29, $0.002 per sample taskGPT-5.5 (OpenAI): index 160.19, $0.11 per sample taskClaude Opus 4.8 (Anthropic): index 159.91, $0.10 per sample taskDeepSeek V4.1 Flash 1017 (DeepSeek): index 159.4, $0.0027 per sample taskGemini 3.8 Flash (Google): index 158.83, $0.017 per sample taskGemini 3.7 Flash (Google): index 158.38, $0.015 per sample taskMuse Spark 1.3 (Meta): index 158.25, $0.021 per sample taskGPT-5.4 (OpenAI): index 157.87, $0.055 per sample taskGrok 4.6 (xAI): index 157.82, $0.032 per sample taskKimi K3 (MoonshotAI): index 158.6, $0.034 per sample taskKimi K3Scatter chart of 29 models, capability against the price of a sample task. Best value at its price: Claude Haiku 5.5, GPT-6.1 Terra, Gemini 4 Carbon, GPT-6.2 Astra.155160165170175$0.001$0.01$0.10$1.00Price of the sample task (log scale)Epoch Capabilities IndexGPT-6.2 Astra (OpenAI): index 174.17, $0.16 per sample taskClaude Fable 5.2 (Anthropic): index 173.5, $0.16 per sample taskGemini 4 Carbon (Google): index 169.78, $0.04 per sample taskGPT-6 Astra (OpenAI): index 169.2, $0.20 per sample taskClaude Opus 5.5 (Anthropic): index 168.7, $0.08 per sample taskGPT-6.1 Sol (OpenAI): index 168.63, $0.04 per sample taskGPT-6.1 Terra (OpenAI): index 167.72, $0.02 per sample taskClaude Fable 5.1 (Anthropic): index 167.09, $0.20 per sample taskClaude Sonnet 5.5 (Anthropic): index 166.59, $0.04 per sample taskGemini 4 Argon (Google): index 166.57, $0.04 per sample taskClaude Opus 5 (Anthropic): index 164.74, $0.10 per sample taskClaude Fable 5 (Anthropic): index 164.21, $0.20 per sample taskGPT-6 Sol (OpenAI): index 163.82, $0.04 per sample taskGPT-5.6 Sol (OpenAI): index 163.05, $0.08 per sample taskGPT-5.5 Pro (OpenAI): index 162.77, $0.66 per sample taskMuse Wonder (Meta): index 161.74, $0.08 per sample taskKimi K3.1 Pro (MoonshotAI): index 161.36, $0.02 per sample taskGPT-5.6 Terra (OpenAI): index 160.78, $0.044 per sample taskGPT-5.4 Pro (OpenAI): index 160.32, $0.66 per sample taskClaude Haiku 5.5 (Anthropic): index 160.29, $0.002 per sample taskGPT-5.5 (OpenAI): index 160.19, $0.11 per sample taskClaude Opus 4.8 (Anthropic): index 159.91, $0.10 per sample taskDeepSeek V4.1 Flash 1017 (DeepSeek): index 159.4, $0.0027 per sample taskGemini 3.8 Flash (Google): index 158.83, $0.017 per sample taskGemini 3.7 Flash (Google): index 158.38, $0.015 per sample taskMuse Spark 1.3 (Meta): index 158.25, $0.021 per sample taskGPT-5.4 (OpenAI): index 157.87, $0.055 per sample taskGrok 4.6 (xAI): index 157.82, $0.032 per sample taskKimi K3 (MoonshotAI): index 158.6, $0.034 per sample taskKimi K3

1 model on the index has no OpenRouter price and is not plotted.

Show the numbers
Each model's capability index and sample-task price
ModelEpoch Capabilities IndexSample taskBest value
GPT-6.2 AstraOpenAI174.17$0.16Yes
Claude Fable 5.2Anthropic173.5$0.16
Gemini 4 CarbonGoogle169.78$0.04Yes
GPT-6 AstraOpenAI169.2$0.20
Claude Opus 5.5Anthropic168.7$0.08
GPT-6.1 SolOpenAI168.63$0.04
GPT-6.1 TerraOpenAI167.72$0.02Yes
Claude Fable 5.1Anthropic167.09$0.20
Claude Sonnet 5.5Anthropic166.59$0.04
Gemini 4 ArgonGoogle166.57$0.04
Claude Opus 5Anthropic164.74$0.10
Claude Fable 5Anthropic164.21$0.20
GPT-6 SolOpenAI163.82$0.04
GPT-5.6 SolOpenAI163.05$0.08
GPT-5.5 ProOpenAI162.77$0.66
Muse WonderMeta161.74$0.08
Kimi K3.1 ProMoonshotAI161.36$0.02
GPT-5.6 TerraOpenAI160.78$0.044
GPT-5.4 ProOpenAI160.32$0.66
Claude Haiku 5.5Anthropic160.29$0.002Yes
GPT-5.5OpenAI160.19$0.11
Claude Opus 4.8Anthropic159.91$0.10
DeepSeek V4.1 Flash 1017DeepSeek159.4$0.0027
Gemini 3.8 FlashGoogle158.83$0.017
Kimi K3MoonshotAI158.6$0.034
Gemini 3.7 FlashGoogle158.38$0.015
Muse Spark 1.3Meta158.25$0.021
GPT-5.4OpenAI157.87$0.055
Grok 4.6xAI157.82$0.032

Our calculation: Epoch AI's index against the sample-task price (undefined tokens in, undefined out, at each model's OpenRouter list price; the formula is under the price chart).

Source: Epoch AI (CC BY), as of Nov 20, 2026Source

Source: OpenRouter, as of Nov 20, 06:00 UTC

In the Drops

Models

Amazon Bedrock adds in-country India inference for Moonshot's Kimi K3

AWS says a new India profile keeps Kimi K3 processing in its Mumbai and Hyderabad regions. Ultrahuman, quoted by AWS, says prompt caching on Bedrock cut its LLM costs by 80 percent.

Why it matters: An open-weight frontier model is now an option for workloads that must stay in India. Check that your IAM and region policies allow only the India profile.

From Drop #065 · Wed, Nov 4, 2026 · Morning
Models

Cerebras explains how it serves the 3T-parameter Kimi K3 across 59 wafers

Cerebras says each WSE-3T wafer holds 44 GB on chip, so it spreads Kimi K3's layers over 59 wafers and passes only activations between them, running multi-trillion parameter models at 1000-2000+ tokens per second.

Why it matters: A worked example of serving frontier-size open weights from on-chip memory instead of moving weights around. The HN post drew 77 points.

From Drop #045 · Sun, Oct 25, 2026 · Morning