claude-opus-4-8 vs kimi-k2.7-code-highspeed

Pricing, Performance & Features Comparison

Authoranthropic

Context Length1M

Reasoning

Providers1

ReleasedMay 2026

Knowledge Cutoff–

License–

Anthropic’s most capable Opus-tier model, tuned for long-horizon agentic work, knowledge work, and memory.

Input$5

Output$25

Latency (p50)5.1s

Output Limit128K

Function Calling

JSON Mode

InputText, Image

OutputText

in$5out$25cache$0.5write$6.3

Authormoonshot

Context Length262K

Reasoning

Providers1

ReleasedJun 2026

Knowledge Cutoff–

License–

High-speed variant of Kimi K2.7-Code (~180 tokens/second) at double the per-token price.

Input$1.9

Output$8

Latency (p50)4.6s

Output Limit–

Function Calling

JSON Mode

InputText

OutputText

in$1.9out$8cache$0.38–