The Known Good Updated 25 Jul 2026
DeepSeek: DeepSeek V3 vs MoonshotAI: Kimi K2.6 ·DeepSeek vs Moonshot AI

DeepSeek: DeepSeek V3 vs MoonshotAI: Kimi K2.6 — head-to-head comparison

DeepSeek: DeepSeek V3 profile MoonshotAI: Kimi K2.6 profile All comparisons
DeepSeek: DeepSeek V3 — DeepSeek
Quality
#44 of 61
44based on 3 of 6 evaluations
Known Good Index
Speed
#280 of 316
20.9
Output tokens per second
Price
#105 of 340
$0.35
Blended per 1M tokens
In $0.20 · Out $0.80
Cache hit price
#10 of 206
$0.003
per 1M tokens (−99%)
Context
#225 of 345
164k
tokens (~300 pages)
MoonshotAI: Kimi K2.6 — Moonshot AI
Quality
#7 of 61
93based on 3 of 6 evaluations
Known Good Index
Speed
#112 of 316
65.7
Output tokens per second
Price
#211 of 340
$1.16
Blended per 1M tokens
In $0.65 · Out $2.72
Cache hit price
#99 of 206
$0.11
per 1M tokens (−83%)
Context
#172 of 345
262k
tokens (~400 pages)
Summary

On the Known Good Index, DeepSeek: DeepSeek V3 scores 44 and MoonshotAI: Kimi K2.6 scores 93 — a 49-point lead for MoonshotAI: Kimi K2.6. They share 2 published evaluations, with DeepSeek: DeepSeek V3 ahead on 0 and MoonshotAI: Kimi K2.6 ahead on 2.

Blended pricing is $0.35 per 1M tokens for DeepSeek: DeepSeek V3 and $1.16 for MoonshotAI: Kimi K2.6 — DeepSeek: DeepSeek V3 is about 3.3× cheaper.

Throughput is 21 tokens per second for DeepSeek: DeepSeek V3 and 66 for MoonshotAI: Kimi K2.6, with time to first token at 1.1s and 1.1s. DeepSeek: DeepSeek V3 accepts 164k tokens of context against 262k.

DeepSeek: DeepSeek V3 is open weights under Open weights (restricted use); MoonshotAI: Kimi K2.6 is open weights under Open weights (unrestricted). 21 tracked hosts serve one or both, and 3 serve both.

Generated from tracked data on 25 Jul 2026 · how these comparisons are computed →

At a glance
MetricDeepSeek: DeepSeek V3MoonshotAI: Kimi K2.6
Known Good Index 44 based on 3 of 6 evaluations 93 based on 3 of 6 evaluations
Blended $/1M $0.35 $1.16
Output tok/s 21 66
Context window 164k 262k
Providers 3 21
Jump to section

Quality

Known Good Index and the capability indices, side by side

IndicesComponents
Index scores

Each index is min-max normalised across tracked models and scaled 0–100 from published evaluations. We do not run these evaluations.

MetricDeepSeek: DeepSeek V3MoonshotAI: Kimi K2.6
Known Good Index 44 93
Mathematics Index 40 96
Science Index 53 95
Reasoning Index 52 —
Agentic Index — 87
Openness Index 60 60
The Known Good

Quality Breakdown

Every evaluation both models report a published score for

Evaluations both models report

Bar = DeepSeek: DeepSeek V3 · marker = MoonshotAI: Kimi K2.6 · scores ingested from Epoch AI (CC-BY 4.0)

GPQA Diamond57 / 91
MATH / AIME16 / 96

Right-hand figures read DeepSeek: DeepSeek V3 / MoonshotAI: Kimi K2.6. DeepSeek: DeepSeek V3 leads on 0 of 2, MoonshotAI: Kimi K2.6 on 2.

The Known Good

Arena Elo

Human preference rating from blind pairwise votes

Arena ratings

Elo from blind pairwise votes, ingested from Arena. Higher is better.

MetricDeepSeek: DeepSeek V3MoonshotAI: Kimi K2.6
Text Arena 1,333 1,455
Text Arena (style controlled) 1,358 1,461
The Known Good

Price & Cost

Published API pricing by token type

Per token typeCache discount
Published pricing

USD per 1M tokens. Blended is a 3:1 input:output mix. Live from OpenRouter, cross-checked against models.dev. Lower is better.

MetricDeepSeek: DeepSeek V3MoonshotAI: Kimi K2.6
Input $/1M $0.20 $0.65
Output $/1M $0.80 $2.72
Blended $/1M $0.35 $1.16
Cache read $/1M $0.003 $0.109
The Known Good
We publish blended price rather than cost-per-task. Cost-per-task requires token counts from an evaluation suite we do not run. Blended price is computed from published rates and is fully reproducible — see the pricing methodology.

Context Window

Maximum input accepted and maximum output emitted

Context and output limits

Provider-declared limits, from metadata via models.dev and OpenRouter. Higher is better.

MetricDeepSeek: DeepSeek V3MoonshotAI: Kimi K2.6
Context window 164k 262k
Max output tokens 16k 262k
The Known Good

Speed & Latency

Median throughput and time to first token across hosting providers

Output speedTime to first token
Throughput and latency

Medians across all tracked providers, derived from OpenRouter endpoint telemetry.

MetricDeepSeek: DeepSeek V3MoonshotAI: Kimi K2.6
Output tokens/s 21 66
Time to first token 1.1s 1.1s
The Known Good

Providers

Which hosts serve each model, and at what price and speed

21 hosts · 3 serve both
Full provider benchmarks →
Provider DeepSeek: DeepSeek V3 blendedDeepSeek: DeepSeek V3 tok/s MoonshotAI: Kimi K2.6 blendedMoonshotAI: Kimi K2.6 tok/s
StreamLake $0.35 24 $1.54 44
DeepInfra $0.46 20 $1.44 57
Novita $0.62 14 $1.45 43
CoreWeave — — $1.34 156
Together — — $2.02 130
Decart — — $1.35 123
Fireworks — — $1.71 111
ModelRun — — $1.38 99
Crusoe — — $1.40 69
Cloudflare — — $1.71 56
Venice — — $1.44 54
DigitalOcean — — $1.37 53
SiliconFlow — — $1.43 50
Chutes — — $1.37 49
AtlasCloud — — $1.71 44
Parasail — — $1.44 42
Moonshot AI · first party — — $1.71 40
Inceptron — — $1.30 39
Baidu · first party — — $1.16 32
BaseTen — — $1.71 —
Phala — — $1.97 —

Per-host pricing and telemetry from OpenRouter. A dash means that host does not serve that model.

The Known Good
Sources for this page
Pricing & providers — OpenRouter Metadata — models.dev Benchmarks — Epoch AI (CC-BY 4.0)Elo — Arena Full attribution →