Skip to main content
Launching the open-source gateway

Hy3 preview vs Kimi K2.5

SpecHy3 previewKimi K2.5
Respan Index28.7 (#87)29.1 (#85)
Blended price per 1M$0.285$1
Context window256K256K
Input / output per 1M$0.18 / $0.6$0.5 / $2.5
Max output--
Inputtexttext, vision, video
Released2026-04-232026-01-27

Respan Index by category

Each category is scored 0 to 100, and how far ahead the winner is. Leads under a point are close to even.

CategoryHy3 previewKimi K2.5
Coding11.6 Hy3 preview9.7 Kimi K2.5
Agents36.8 Hy3 preview31.5 Kimi K2.5
Reasoning27.2* Hy3 preview30.6 Kimi K2.5
Knowledge & language23.9* Hy3 preview33.9 Kimi K2.5
Human preference49.3 Hy3 preview51.6 Kimi K2.5

* Estimated: no published scores in that category. How the Respan Index works

Benchmarks both report

Kimi K2.5 leads on 4 of the 5 benchmarks both report. Higher is better. Each score links to where it was published.

BenchmarkHy3 previewKimi K2.5
LMArena Elo14091450
LMArena WebDev (LMArena)13571436
BrowseComp67.1%60.6%
GPQA-Diamond87.2%87.6%
SWE-bench Verified74.4%76.8%

FAQ

Is Hy3 preview better than Kimi K2.5?

Kimi K2.5 scores 29.1 on the Respan Index (#85), ahead of Hy3 preview at 28.7 (#87). Kimi K2.5 leads on 4 of the 5 benchmarks both report.

Which is cheaper, Hy3 preview or Kimi K2.5?

Hy3 preview lists at $0.18 input and $0.6 output per 1M tokens; Kimi K2.5 at $0.5 and $2.5. Hy3 preview is 3.5x cheaper per token at list price.

Which has the larger context window?

Both accept 256K tokens of context.

Compare other models

vs

Built for AI agents.
Break less.
Ship more.