Benchmark Heaven
Price & provider filters · adjusted costs
Global
Applies to price views & model offers; benchmark evidence stays unfiltered
← All models

Kimi K3

open weights★ featured
Moonshot AI · 24 recorded token offers

Top 5 cheapest providers (Adjusted $/task)

Within the active global provider, residency and confidentiality filters.

#ProviderPlatformRaw input $/1MRaw output $/1MAdjusted $/task
1Sail ResearchOpenRouter$2.60$13.00
2TogetherOpenRouter$3.00$15.00
3RelaceOpenRouter$2.40$12.00
4MorphOpenRouter$2.50$14.00
5ModalOpenRouter$3.00$15.00

Benchmarks

Composite92.4
Composite evidence1/5
AA Coding · unversioned snapshot 2026-09-10
AA Coding Agent · v1.4 · 2026-09-0962.6
AA Intelligence · unversioned snapshot 2026-09-10
DesignArena Frontend · unversioned snapshot 2026-09-10
DesignArena Full-Stack · unversioned snapshot 2026-09-10
Output speed (t/s)

MODEL EVIDENCE

Benchmark sheet

5 / 73 registered benchmark versions covered · 5 observations including existing index snapshots.

Compare this model ↗
Composite retains Coding Agent v1.4 · source 2026-09-09

The five Composite inputs remain unchanged. Its Coding Agent input is the median across complete harness results in the retained v1.4 snapshot from 2026-09-09. Artificial Analysis now publishes v1.5, with different components. Explore current v1.5 separately. Dated snapshot labels on other indices identify unversioned source captures, not a verified semantic version.

Profile signals

Insufficient evidence to evaluate · 0 eligible benchmark families

How flags are calculated

Heuristic screen, not statistical significance: benchmark families are correlated and source uncertainty is unknown. Peer evidence needs ≥ 20 independently measured matched configurations from ≥ 10 distinct model families. A flag needs a directed population z-score of magnitude ≥ 1.5 and a gap of ≥ 1.5 from the leave-one-benchmark-family-out mean z in the same direction, over ≥ 5 other benchmark families.

Protocol-compatible measured/vendor divergences

No verified protocol-compatible vendor/measured pair is available for this model; agreement cannot be assessed.

Coding

Coding: every observation for this exact model configuration
Benchmark / versionEvaluation groupScore / provenance
Artificial Analysis Coding Agent Index v1.4

Version 1.4

Kimi Code CLI

The retained AA Coding Agent Index measures coding-agent systems using the earlier three-component implementation.

0.62639 fractionmeasuredobserved 2026-09-09artificialanalysis.ai
Evidence
Axis: Artificial Analysis Coding Agent Index v1.4 · 1.4 · Kimi Code CLI
Exact value: 0.6263880112748013 fraction
Observed: 2026-09-09 · publication date: not recorded
Observation id: aa-coding:1.4:52
Artificial Analysis Coding Agent Index v1.5

Version 1.5

Kimi Code CLI

Measures coding-agent systems on DeepSWE v1.1, Terminal-Bench 4.0 and SWE-Atlas-QnA.

0.51926 fractionmeasuredobserved 2026-09-10artificialanalysis.ai
Evidence
Axis: Artificial Analysis Coding Agent Index v1.5 · 1.5 · Kimi Code CLI
Exact value: 0.519259105470924 fraction
Observed: 2026-09-10 · publication date: not recorded
Observation id: aa-coding:1.5:55b029d6971b4d378335bb168eb98b3b

Roleplay

Roleplay: every observation for this exact model configuration
Benchmark / versionEvaluation groupScore / provenance
Japanese-RP-Bench v2 (tegnike fork)

Version 2

Published board

Fork of Japanese-RP-Bench adding v2 metrics for role fidelity, conversation quality, persona stability, robustness, and recovery in Japanese roleplay.

89.59 pointsmeasuredobserved 2026-09-10raw.githubusercontent.com
Evidence
Axis: Japanese-RP-Bench v2 (tegnike fork) · 2 · Published board
Exact value: 89.59 points
Observed: 2026-09-10T21:53:18Z · publication date: not recorded
Observation id: public:9f37d75c0b2c80f0c1c0d4c4

Writing

Writing: every observation for this exact model configuration
Benchmark / versionEvaluation groupScore / provenance
LLM Creative Story-Writing Benchmark

Version snapshot-2026-09-10

Published board

Pairwise comparison of short stories written to the same constrained creative briefs, with LLM evaluator choices combined into a relative comparison score.

2.7 pointsmeasuredobserved 2026-09-10raw.githubusercontent.com
Evidence
Axis: LLM Creative Story-Writing Benchmark · snapshot-2026-09-10 · Published board
Exact value: 2.7 points
Observed: 2026-09-10T21:53:00Z · publication date: not recorded
Observation id: public:d41293fa05950343ccf0be08
The Slop Index

Version snapshot-2026-09-10

Published board

Leaderboard scoring how much each model's writing reads like chatbot slop versus pre-AI human writing, blending live arena votes with four mechanical axes.

26.6 pointsmeasuredobserved 2026-09-10theslopindex.com
Evidence
Axis: The Slop Index · snapshot-2026-09-10 · Published board
Exact value: 26.6 points
Observed: 2026-09-10T23:12:30.991903+00:00 · publication date: not recorded
Observation id: public:c7dad535e7db9249f6eddd35
Missing coverage · 72 benchmark versions

No result does not mean a zero, or that the model was never tested. Collection failures and disputed versions retain their distinct status.

Variants / reasoning settings

VariantCoding · snapshot 2026-09-10Intelligence · snapshot 2026-09-10
Kimi K3 (max)76.243.8
Kimi K3 (low)72.034.5
Kimi K3

GitHub Copilot

Current usage-based billing · model token cost is converted to AI Credits at 1 credit = $0.01.

Input $3.00 / 1M
Cached input $0.300 / 1M
Output $15.00 / 1M
Status GA

Token offers by platform — Adjusted $/task

Adjusted costs are modeled USD per task: AA output tokens × OpenRouter usage I/O (Chutes global fallback). These are general usage and benchmark proxies for coding-agent work. Missing AA data assumes 1,000 output tokens/task; unknown cache hit assumes 0%; unmeasured additional cache writes assume 0 tokens. Click any underlined price for exact inputs, dates and assumptions. Raw list prices use the selected fixed input/output blend, in USD per million tokens.

22 offers within the active global filters; “—” means the catalog is active but no public token price is available.

OpenRouter (17)

Sail Researchglobalsail-research/fp4$2.60 raw in $/1M$13.00 raw out $/1M
Togetherglobaltogether$3.00 raw in $/1M$15.00 raw out $/1M
Morphglobalmorph/fp4$2.50 raw in $/1M$14.00 raw out $/1M
Modalglobalmodal/mxfp4$3.00 raw in $/1M$15.00 raw out $/1M
Fireworksglobalfireworks$3.00 raw in $/1M$15.00 raw out $/1M
DigitalOceanglobaldigitalocean$2.55 raw in $/1M$12.95 raw out $/1M
BaseTenglobalbaseten/fp8$3.00 raw in $/1M$15.00 raw out $/1M
PhalaglobalphalaTEE$3.00 raw in $/1M$15.00 raw out $/1M
Fireworksglobalfireworks/us$3.30 raw in $/1M$16.50 raw out $/1M
DeepInfraglobaldeepinfra/bf16$2.85 raw in $/1M$14.25 raw out $/1M
Makoraglobalmakora$2.55 raw in $/1M$12.75 raw out $/1M
Waferglobalwafer$3.00 raw in $/1M$12.75 raw out $/1M
Parasailglobalparasail/fp4$3.00 raw in $/1M$15.00 raw out $/1M
Morphglobalmorph/fast$6.00 raw in $/1M$22.50 raw out $/1M
Chutesglobalchutes/mxfp4$3.00 raw in $/1M$15.00 raw out $/1M
Fireworksglobalfireworks/fast$4.50 raw in $/1M$22.50 raw out $/1M
Relaceglobalrelace/fp4$2.40 raw in $/1M$12.00 raw out $/1M

Nebius (1)

NebiuseuEU$3.00 raw in $/1M$15.00 raw out $/1M

TensorX (1)

TensorXeuEU$3.00 raw in $/1M$15.00 raw out $/1M

Chutes (1)

ChutesglobalTEE$3.00 raw in $/1M$15.00 raw out $/1M

Azure AI Foundry (2)

Azure AI Foundryus$3.30 raw in $/1M$16.50 raw out $/1M
Azure AI Foundryusfireworks$3.30 raw in $/1M$16.50 raw out $/1M
Benchmark Heaven — Model benchmarks & costs