/
Recipes
Search
Search models or providers...
⌘K
Browse
Docs
GitHub
Preview
of
vllm-project/vllm#60762
at
389ffeb
.
Providers
Arcee AI
AutoTrust
Ernie (Baidu)
BharatGen
Black Forest Labs
Boson AI
Seed (ByteDance)
DeepSeek
Dots
Fish Audio
Google
IBM Granite
IFM
inclusionAI
IndexTeam
InternLM
JetBrains
Jina AI
Lightricks
Liquid AI
LongCat (Meituan)
Meta
Muse (Meta)
Microsoft
MindLab Research
MiniMax
Mistral AI
Moonshot AI
Kimi-K3
Kimi-K2.7-Code
Kimi-K2.6
Kimi-K2.5
Kimi-K2-Thinking
Kimi-Linear-48B-A3B-Instruct
Kimi-K2-Instruct
NVIDIA
OpenAI
MiniCPM (OpenBMB)
InternVL (OpenGVLab)
OpenMOSS
PaddlePaddle
Preferred Networks
Poolside
Qwen
RecursiveSI
Reflection AI
Stability AI
StepFun
Hunyuan (Tencent)
Thinking Machines Lab
vLLM Semantic Router
Wan (Alibaba)
MiMo (Xiaomi)
GLM (Z-AI)
Moonshot AI
moonshotai
·
7 recipes
HuggingFace
Multimodal
4
Kimi-K3
2.8T / 16 experts/token + shared (of 896 routed)
moe
MXFP4
1680G
NVFP4
1650G
INT4
1711G
FP8
3360G
v0.29.0+
→
Kimi-K2.7-Code
1T / 32B
moe
INT4
714G
v0.19.1+
→
Kimi-K2.6
1T / 32B
moe
INT4
714G
NVFP4
600G
FP8
1200G
NVFP4
600G
v0.25.0+
→
Kimi-K2.5
1T / 32B
moe
INT4
714G
NVFP4
600G
MXFP4
669G
v0.19.1+
→
Text
3
Kimi-K2-Thinking
1T / 32B
moe
INT4
600G
NVFP4
600G
FP8
1200G
v0.12.0+
→
Kimi-Linear-48B-A3B-Instruct
48B / 3B
moe
BF16
115G
v0.11.2+
→
Kimi-K2-Instruct
1T / 32B
moe
FP8
1200G
v0.12.0+
→
/Recipes/Recipes