Kimi K3
Moonshot's flagship — long-horizon coding and knowledge work, 2.8T parameters.
Kimi K3 is Moonshot's flagship: 2.8 trillion parameters, native vision, a 1M-token context, effort up to max. Moonshot positions it for long-horizon coding and knowledge work. TeamCLI runs it today as a Claude Code endpoint seat; the native Kimi CLI is listed but has no launch profile yet.
What it is good at
- Long-horizon coding and knowledge work — Moonshot's own positioning
- Native vision with a 1M-token context
- Cached input at $0.30 per MTok — a tenth of fresh input
What it is less good at
- No benchmarks published by Moonshot
- No native-CLI launch profile yet — the endpoint route is the way in today
Reaching it with TeamCLI
TeamCLI runs Kimi K3 as Kimi K3:
Ways to reach it
| Route | Spends | Status |
|---|---|---|
| Claude Code with a provider endpoint | A Moonshot (Kimi platform) API key. | Verified |
| Kimi Code CLI | A Kimi Code membership or a Kimi platform API key. | Listed — not yet verified |
| The API directly | Moonshot's API billing. | Listed — not yet verified |
The seat line
Each route maps to a TeamCLI seat line — the same fields, written generically. No machine paths, no launcher script of anybody's, no team's own seat names.
| Route | cli | vendor | model | version | display | launch |
|---|---|---|---|---|---|---|
| Claude Code with a provider endpoint | claude-code | moonshot | Kimi | K3 | Kimi K3 | my-lab-claude |
A seat with model_from: launcher leaves the running model to what the launch runs — the provider's launcher, or Jules — so team doctor reads the model from the seat rather than from the file.
Known quirks
- The native Kimi Code CLI exists but is not a TeamCLI seat kind yet, so it has no launch profile.
Versions
One block per version, the current one first, each with its own anchor. The comparison table puts the costs and the family's benchmark columns side by side; the values are verbatim, and every one is sourced and dated in its version's section below and in the Sources list. A dash means the source publishes no figure, and nothing here is estimated.
No independent intelligence or speed index is republished here: Artificial Analysis measures both across labs, and its terms allow the link but not the copy — its figures live on its own pages.
| Version | Released | Context | Input price | Output price |
|---|---|---|---|---|
| Kimi K3 — current | — | 1,048,576 tokens | $3.00 / MTok | $15.00 / MTok |
Kimi K3 — current
- Model id:
kimi-k3
Abilities
- Released: —
- Context window: 1,048,576 tokens — Moonshot — Kimi pricing (read 2026-10-19).
- Max output: —
- Output speed: —
- Latency: —
Costs
- Input price: $3.00 / MTok — Moonshot — Kimi pricing (read 2026-10-19). Cached input $0.30; cache writes $3.00 (5m) / $6.00 (1h)
- Output price: $15.00 / MTok — Moonshot — Kimi pricing (read 2026-10-19).
This is the version the seat lines above run.
Sources
- Moonshot — Kimi pricing — read 2026-10-19
Figures refreshed 2026-10-19; benchmarks are vendor-reported. Prices are per million tokens unless a figure says otherwise.