difference

compare_arrowsModel decision desk

Compare AI models side by side

Select 2 to 4 models and compare official pricing, context, capabilities and measured relay availability in one view.

199Models

2–4Side by side

38Relay records

Selected models

Up to 4 models

openai

GPT-5.6 Sol

gpt-5.6-sol

Estimated cost$8for this usage

anthropic

Claude Opus 5

claude-opus-5

Estimated cost$10for this usage

Moonshot

Kimi-K3

kimi-k3

Estimated cost¥40for this usage

Usage scenario

Compare costs under the same workload

Input tokensOutput tokensCalls

Estimated totals above update instantly

Comparison matrix

Price and capability data comes from the public model price directory. Relay counts, latency and online rate are matched by exact model ID from the relay ranking.

Metric

openaiGPT-5.6 Sol

anthropicClaude Opus 5

MoonshotKimi-K3

apartmentProvider

openai

anthropic

Moonshot

deployed_codeModel ID

gpt-5.6-sol

claude-opus-5

kimi-k3

loginInput / 1M tokens

¥20

logoutOutput / 1M tokens

$20

$25

¥100

cachedCache read / 1M

$0.4

$0.5

¥2

saveCache write / 1M

$6.25

—

data_arrayContext window

1.1M

psychologyReasoning

checkSupported

buildTool calling

checkSupported

visibilityVision input

checkSupported

hubAvailable relays

No data

speedBest measured latency

2175 ms

No data

monitor_heartBest online rate

100%

No data