The Known Good Updated 4 Sep 2026 Subscribe
DeepSeek: DeepSeek V4 Pro 0423 vs MoonshotAI: Kimi K2.6 ·DeepSeek vs Moonshot AI

DeepSeek: DeepSeek V4 Pro 0423 vs MoonshotAI: Kimi K2.6 — head-to-head comparison

DeepSeek: DeepSeek V4 Pro 0423 profile MoonshotAI: Kimi K2.6 profile All comparisons
DeepSeek: DeepSeek V4 Pro 0423 — DeepSeek
Quality
#4 of 65
93based on 3 of 7 evaluations
Known Good Index
Speed
#201 of 338
42.2
Output tokens per second
Price
#263 of 442
$1.30
Blended per 1M tokens
In $1.04 · Out $2.08
Cache hit price
#151 of 324
$0.087
per 1M tokens (−92%)
Context
#43 of 476
1.05M
tokens (~1,600 pages)
MoonshotAI: Kimi K2.6 — Moonshot AI
Quality
#8 of 65
92based on 3 of 7 evaluations
Known Good Index
Speed
#154 of 338
52.1
Output tokens per second
Price
#290 of 442
$1.71
Blended per 1M tokens
In $0.95 · Out $4.00
Cache hit price
#211 of 324
$0.16
per 1M tokens (−83%)
Context
#264 of 476
262k
tokens (~400 pages)
Summary

On the Known Good Index, DeepSeek: DeepSeek V4 Pro 0423 scores 93 and MoonshotAI: Kimi K2.6 scores 92, which puts them level. They share 3 published evaluations, with DeepSeek: DeepSeek V4 Pro 0423 ahead on 3 and MoonshotAI: Kimi K2.6 ahead on 0.

Blended pricing is $1.30 per 1M tokens for DeepSeek: DeepSeek V4 Pro 0423 and $1.71 for MoonshotAI: Kimi K2.6 — DeepSeek: DeepSeek V4 Pro 0423 is about 1.3× cheaper.

Throughput is 42 tokens per second for DeepSeek: DeepSeek V4 Pro 0423 and 52 for MoonshotAI: Kimi K2.6, with time to first token at 1.5s and 1.8s. DeepSeek: DeepSeek V4 Pro 0423 accepts 1.05M tokens of context against 262k.

DeepSeek: DeepSeek V4 Pro 0423 is open weights under Open weights (unrestricted); MoonshotAI: Kimi K2.6 is open weights under Open weights (unrestricted). 29 tracked hosts serve one or both, and 15 serve both.

Generated from tracked data on 4 Sep 2026 · how these comparisons are computed →

At a glance
MetricDeepSeek: DeepSeek V4 Pro 0423MoonshotAI: Kimi K2.6
Known Good Index 93 based on 3 of 7 evaluations 92 based on 3 of 7 evaluations
Blended $/1M $1.30 $1.71
Output tok/s 42 52
Context window 1.05M 262k
Providers 21 23
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 V4 Pro 0423MoonshotAI: Kimi K2.6
Known Good Index 93 92
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 V4 Pro 0423 · marker = MoonshotAI: Kimi K2.6 · scores ingested from Epoch AI (CC-BY 4.0)

GPQA Diamond91 / 91
Mock AIME 2024–2597 / 96
SWE-bench Verified78 / 77

Right-hand figures read DeepSeek: DeepSeek V4 Pro 0423 / MoonshotAI: Kimi K2.6. DeepSeek: DeepSeek V4 Pro 0423 leads on 3 of 3, MoonshotAI: Kimi K2.6 on 0.

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 V4 Pro 0423MoonshotAI: Kimi K2.6
Text Arena 1,451 1,455
Text Arena (style controlled) 1,458 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 V4 Pro 0423MoonshotAI: Kimi K2.6
Input $/1M $1.04 $0.95
Output $/1M $2.08 $4.00
Blended $/1M $1.30 $1.71
Cache read $/1M $0.087 $0.160
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 V4 Pro 0423MoonshotAI: Kimi K2.6
Context window 1.05M 262k
Max output tokens 384k 236k
The Known Good

Speed & Latency

Mean throughput and time to first token across hosting providers, over successful measurements from the last 48 hours

Throughput and latency

Means across all tracked providers over successful measurements from the last 48 hours, derived from OpenRouter endpoint telemetry.

MetricDeepSeek: DeepSeek V4 Pro 0423MoonshotAI: Kimi K2.6
Output tokens/s 42 52
Time to first token 1.5s 1.8s
The Known Good

Providers

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

29 hosts · 15 serve both
Full provider benchmarks →
Provider DeepSeek: DeepSeek V4 Pro 0423 blendedDeepSeek: DeepSeek V4 Pro 0423 tok/s MoonshotAI: Kimi K2.6 blendedMoonshotAI: Kimi K2.6 tok/s
BaseTen $2.17 66 $1.71 —
CoreWeave $1.50 62 $1.34 140
Novita $2.00 58 $1.45 28
Azure · first party $2.39 57 — —
Alibaba · first party $1.77 54 — —
Parasail $2.17 52 $1.44 78
NextBit $2.15 49 — —
Venice $2.06 49 $1.44 8
Baidu · first party $2.11 48 $1.71 26
AtlasCloud $2.10 43 $1.71 22
DeepInfra $1.62 43 $1.44 7
Together $2.17 39 $2.02 119
SiliconFlow $1.91 34 $1.43 23
DeepSeek · first party $0.99 31 — —
GMICloud $1.30 31 $1.54 10
StreamLake $1.30 30 $1.08 33
DigitalOcean $1.09 28 $1.71 41
Wafer $1.50 28 — —
Ionstream $1.41 26 — —
Cloudflare $1.50 12 $1.71 39
Fireworks $1.20 — $1.71 38
ModelRun — — $1.38 103
Decart — — $1.00 65
Phala — — $1.97 57
Inceptron — — $1.24 38
Sail Research — — $1.75 33
Moonshot AI · first party — — $1.71 26
Chutes — — $1.28 23
Crusoe — — $1.40 9

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 →