Kimi K3.1 Pro: API-Verfügbarkeit
Bestätigt24 Beleglinks
Englisches Original

Moonshot
USD / 1 Mio. Token
| Eingabe | Ausgabe | Cache-Eingabe |
|---|---|---|
|
|
|
| Modell | Beleglinks | Preis Ein-/Ausgabe |
|---|---|---|
moonshotai/kimi-k3.1-prostandard | openrouter | $2,00 / $10,00 |
moonshotai/kimi-k3.1-pro:batchbatch | openrouter | $0,70 / $6,00 |
moonshotai/kimi-k3.1-prostandard | vercel-gateway | $2,00 / $10,00 |
kimi-k3.1-prostandard · Text | opencode-go | $2,00 / $10,00 |
Kimi K3.1 Pro is NOT callable. build.nvidia.com publishes the model, but its
The form showed "Для Strix: openai/kimi-k3.1-pro" under a model with no such
- With Kimi-K3.1-Pro and the recipe's DSpark draft (`kv_cache_dtype: fp8`) on 8x MI325X, draft acceptance at 3 speculative tokens no longer collapses after a batch
- Kimi K3.1 Pro (`moonshotai/kimi-k3.1-pro`) is now a Token Pass model.
feat(devin): add claude-fable-5-2 and kimi-k3-1-pro to default catalog
Models: new `hatch_model_roles` roles `inspiration` and `content`, seeded `kimi-k3.1-pro`, resolved per session role.
Models: Qwen3.8-2.4T, Qwen3.8-27B, Gemma 4 31B, DeepSeek-V4.1-Flash, DeepSeek-V4-Pro, Kimi-K3.1-Pro, GLM-5.3, MiniMax-M3.1-Flash, Mistral-Large-4, Muse-Spark-Open, Nemotron-3.5-Super, Hy4-preview; small: LFM3-1.7B, Ling-3.0-tiny, granite-4.2-3b, granite-4.2-8b, Qwen3.5-9B, LFM2.5-24B, Nemotron-3.5-Lightning, Gemma 4
[`bcb0152`](https://redirect.github.com/ggml-org/llama.cpp/commit/bcb015287) model: add Kimi-K3.1-Pro text model ([#​31779](https://redirect.github.com/ggml-org/llama.cpp/issues/31779))
documents that missing field as HTTP 400. The `kimi-k3.1-pro[-highspeed]` direct-platform spellings did replay at default effort via the generic `kimi-` prefix heuristic, but were still
| Kimi-K3.1-Pro attention shape (96 heads, 4096 tokens, top-k 2048), pairs, 1x H100 | first call about 7 s (compile), then 12 ms forward and
<li>ai-sdk-provider: add kimi-k3-1-pro to known model ids by <a href="https://github.com/andrelandgraf"><code>@andrelandgraf</code></a> in <a href="https://redirect.github.com/neondatabase/neon-pkgs/pull/1091">neondatabase/neon-pkgs#1091</a></li>
<li>ai-sdk-provider: add kimi-k3-1-pro to known model ids by <a href="https://github.com/andrelandgraf"><code>@andrelandgraf</code></a> in <a href="https://redirect.github.com/neondatabase/neon-pkgs/pull/1091">neondatabase/neon-pkgs#1091</a></li>
Kimi-K3.1-Pro on 8x MI355X (TP8, KVP8, FP8 KV cache, 3 DSpark drafts, the gfx950 stack through #3355), InferenceX's acceptance-length gate: SPEED-Bench coding prompts, 80 requests
rows showed prices. `cn:deepseek-v4.1-flash` came from OpenRouter (0.3/1.2). `cn:kimi-k3.1-pro`, `cn:glm-5.3-flash`, `zd/gpt-6-sol` and `zd/gemini-4-argon` came from models.dev.
- created: `/home/runner/_work/models.dev/models.dev/providers/requesty/models/kimi-k3.1-pro@eu.toml`
kimi-k3.1-proKimi K3.1 Pro ist ein Modell von Moonshot mit offenen Modellgewichten, veröffentlicht am 2026-10-23. Das Kontextfenster umfasst 1.048.576 Tokens. Pro Million Tokens kostet die Eingabe $2,00 und die Ausgabe $10,00.
Das Kontextfenster umfasst 1.048.576 Tokens.
Pro Million Tokens kostet die Eingabe $2,00 und die Ausgabe $10,00. Cache-Eingabe kostet $0,20 pro Million.
Ja. Moonshot veröffentlicht die Modellgewichte.
Am 2026-10-23.