The Known Good Updated 29 Jul 2026
DeepSeek: DeepSeek V3 0324 vs MoonshotAI: Kimi K2.6 ·DeepSeek vs Moonshot AI

DeepSeek: DeepSeek V3 0324 vs MoonshotAI: Kimi K2.6 — head-to-head comparison

DeepSeek: DeepSeek V3 0324 profile MoonshotAI: Kimi K2.6 profile All comparisons
DeepSeek: DeepSeek V3 0324 — DeepSeek
Quality
#38 of 62
64based on 4 of 7 evaluations
Known Good Index
Speed
#272 of 317
24.6
Output tokens per second
Price
#140 of 370
$0.48
Blended per 1M tokens
In $0.27 · Out $1.12
Cache hit price
#141 of 238
$0.14
per 1M tokens (−50%)
Context
#254 of 375
164k
tokens (~300 pages)
MoonshotAI: Kimi K2.6 — Moonshot AI
Quality
#7 of 62
93based on 3 of 7 evaluations
Known Good Index
Speed
#122 of 317
58.8
Output tokens per second
Price
#225 of 370
$1.16
Blended per 1M tokens
In $0.65 · Out $2.72
Cache hit price
#119 of 238
$0.11
per 1M tokens (−83%)
Context
#199 of 375
262k
tokens (~400 pages)
Summary

On the Known Good Index, DeepSeek: DeepSeek V3 0324 scores 64 and MoonshotAI: Kimi K2.6 scores 93 — a 29-point lead for MoonshotAI: Kimi K2.6. They share 2 published evaluations, with DeepSeek: DeepSeek V3 0324 ahead on 0 and MoonshotAI: Kimi K2.6 ahead on 2.

Blended pricing is $0.48 per 1M tokens for DeepSeek: DeepSeek V3 0324 and $1.16 for MoonshotAI: Kimi K2.6 — DeepSeek: DeepSeek V3 0324 is about 2.4× cheaper.

Throughput is 25 tokens per second for DeepSeek: DeepSeek V3 0324 and 59 for MoonshotAI: Kimi K2.6, with time to first token at 1.2s and 1.4s. DeepSeek: DeepSeek V3 0324 accepts 164k tokens of context against 262k.

DeepSeek: DeepSeek V3 0324 is open weights; MoonshotAI: Kimi K2.6 is open weights under Open weights (unrestricted). 22 tracked hosts serve one or both, and 4 serve both.

Generated from tracked data on 29 Jul 2026 · how these comparisons are computed →

At a glance
MetricDeepSeek: DeepSeek V3 0324MoonshotAI: Kimi K2.6
Known Good Index 64 based on 4 of 7 evaluations 93 based on 3 of 7 evaluations
Blended $/1M $0.48 $1.16
Output tok/s 25 59
Context window 164k 262k
Providers 4 22
Jump to section

Quality

Known Good Index and the capability indices, side by side

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 V3 0324MoonshotAI: Kimi K2.6
Known Good Index 64 93
Mathematics Index 57 96
Reasoning Index 65 —
Agentic Index — 87
Openness Index 100 60
The Known Good

Quality Breakdown

Every evaluation both models report a published score for

Evaluations both models report

Bar = DeepSeek: DeepSeek V3 0324 · marker = MoonshotAI: Kimi K2.6 · scores ingested from Epoch AI (CC-BY 4.0)

GPQA Diamond68 / 91
Mock AIME 2024–2538 / 96

Right-hand figures read DeepSeek: DeepSeek V3 0324 / MoonshotAI: Kimi K2.6. DeepSeek: DeepSeek V3 0324 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 V3 0324MoonshotAI: Kimi K2.6
Text Arena 1,375 1,455
Text Arena (style controlled) 1,395 1,461
The Known Good

Price & Cost

Published API pricing by token type

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 V3 0324MoonshotAI: Kimi K2.6
Input $/1M $0.27 $0.65
Output $/1M $1.12 $2.72
Blended $/1M $0.48 $1.16
Cache read $/1M $0.135 $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 V3 0324MoonshotAI: Kimi K2.6
Context window 164k 262k
Max output tokens 66k 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 V3 0324MoonshotAI: Kimi K2.6
Output tokens/s 25 59
Time to first token 1.2s 1.4s
The Known Good

Providers

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

22 hosts · 4 serve both
Full provider benchmarks →
Provider DeepSeek: DeepSeek V3 0324 blendedDeepSeek: DeepSeek V3 0324 tok/s MoonshotAI: Kimi K2.6 blendedMoonshotAI: Kimi K2.6 tok/s
Crusoe $0.75 33 $1.40 50
Novita $0.48 29 $1.45 31
DeepInfra $0.41 21 $1.44 38
SiliconFlow $0.44 17 $1.43 34
Together — — $2.02 138
CoreWeave — — $1.34 135
Parasail — — $1.44 128
BaseTen — — $1.71 127
Fireworks — — $1.71 86
Decart — — $1.34 76
ModelRun — — $1.38 76
Cloudflare — — $1.71 56
Inceptron — — $1.30 52
DigitalOcean — — $1.37 51
Phala — — $1.97 42
Baidu · first party — — $1.16 42
Chutes — — $1.37 42
Venice — — $1.44 36
AtlasCloud — — $1.71 34
StreamLake — — $1.54 31
Moonshot AI · first party — — $1.71 23
Sail Research — — $1.75 6

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 →