WhichAI / Models / Kimi K3

Kimi K3

· verified October 14, 2026

public free paid api open-weights in Generator coding reasoning agents vision long-context

Provider

Statuspublic

AA Intelligence Index43.6 (official, read October 14, 2026)

Released2026-07-16

Context window1M

ModalitiesText + vision

API price $/1M (in / out)$0.58 / $15 (OpenRouter)

Speedn/a

AccessAPI $0.58/$15 per 1M tokens ($0.30 cached) · free kimi.com chat · open weights on Hugging Face (July 27, Modified MIT)

In one line

Moonshot's July flagship: 2.8T-parameter MoE, 1M context, native vision, and the best published BrowseComp score (91.2%). The strongest agentic challenger to the US frontier, and the largest model whose weights you can actually download.

Coding90

Reasoning91

Writing78

Agents & tools94

Category scores (0-100) are WhichAI blended ratings: public category leaderboards plus editorial judgment. Guidance, not official measurements.

Alternatives worth a look

Claude Haiku 5.5 (Anthropic) · AA 43.4API $0.1/$0.5 per 1M tokens on OpenRouter · listed October 7, 2026
Grok 4.6 (xAI) · AA 44.3Released August 12, 2026 · API $2/$6 per 1M tokens ($0.50 cached) · in Grok Build and Cursor · SuperGrok subscription
GLM-5.3 (Z.ai) · AA 44.8Released August 14, 2026 · staged access · open weights NOT released at launch (unlike every previous GLM) · API price not published

Head-to-head: Kimi K3 vs GPT-5.6 Sol · Kimi K3 vs GLM-5.2 · Kimi K3 vs Claude Opus 4.8

Open Kimi's official app ↗ Open in the WhichAI app Not sure? Take the 30-second finder

Artificial Analysis Intelligence Index, official scale (the number published on artificialanalysis.ai), relayed by OpenRouter and refreshed automatically: last read October 14, 2026. 209 models carry a current official measurement. A value marked ~ is converted from an earlier snapshot because no current measurement is published; an unmeasured model is never ranked. Context, price, modality and listing dates come from the public OpenRouter model list (refreshed automatically on the live site; static copy retrieved October 3, 2026) and vendor pages. Missing values are not published or not yet verified - shown as -. Sources are linked inside the app (the Ranking chapter and About > Methodology).