AI 模型库

浏览所有可用 AI 模型,支持多个推理框架

查看更换模型部署指南
模型总数 105
当前结果 105
可用模型 105
显示 1-24 / 105 个模型
第 1 / 5
BAAI
Bge_0.5b_0.5k_b1_2core_1chip
0.5 gguf w4a8 M50 512
文本 向量
HLIECpp HLIEPython HLIELLama

Type: Bge | Size: 0.5 | Quant: w4a8 | Context: 512 | Chip: M50 | Build: 1.4.0 | Status: Stable

BAAI
Bge-Reranker_0.5b_0.5k_b1_2core_1chip
0.5 gguf w4a8 M50 512
文本 向量
HLIECpp HLIEPython HLIELLama

Type: Bge-Reranker | Size: 0.5 | Quant: w4a8 | Context: 512 | Chip: M50 | Build: 1.4.0 | Status: Stable

CogVideoX
Copaw_9b_128k_b1_2core_1chip
9 gguf w4a8 M50 131,072
文本 文本
HLIECpp HLIEPython HLIELLama

Type: Copaw | Size: 9 | Quant: w4a8 | Context: 131072 | Chip: M50 | Build: 1.3.0 | Status: Stable

CogVideoX
Copaw_9b__b1_2core_1chip(448*448)
9 gguf w4a8 M50 视觉
文本 文本
HLIECpp HLIEPython HLIELLama

Type: Copaw | Size: 9 | Quant: w4a8 | Chip: M50 | Build: 1.4.0 | Status: Stable

阿里云
CosyVoice3_0.5b_2k_b1_2core_1chip
0.5 gguf w4a8 M50 2,048
文本 音频
HLIECpp HLIEPython HLIELLama

Type: CosyVoice3 | Size: 0.5 | Quant: w4a8 | Context: 2048 | Chip: M50 | Build: 1.4.0 | Status: Stable

DEEPSEEK
DeepSeek_8b_32k_b1_2core_1chip
8 gguf w4a8 M50 32,768
文本 文本
HLIECpp HLIEPython HLIELLama

Type: DeepSeek | Size: 8 | Quant: w4a8 | Context: 32768 | Chip: M50 | Build: 1.3.0 | Status: Stable

Google
Gemma4_2b_128k_b1_2core_1chip()
2 gguf w4a8 M50 131,072 视觉
文本 文本
HLIECpp HLIEPython HLIELLama

Type: Gemma4 | Size: 2 | Quant: w4a8 | Context: 131072 | Chip: M50 | Build: 1.4.0 | Status: Stable

Google
Gemma4_26b-a4b-mtp_256k_b1_2core_1chip(768*2520)
26 gguf w4a8 M50 262,144 视觉
文本 文本
HLIECpp HLIEPython HLIELLama

Type: Gemma4 | Size: 26 | Quant: w4a8 | Context: 262144 | Chip: M50 | Build: 1.4.0 | Status: Temporary

Google
Gemma4_26b-a4b_128k_b1_2core_1chip
26 gguf w4a8 M50 131,072
文本 文本
HLIECpp HLIEPython HLIELLama

Type: Gemma4 | Size: 26 | Quant: w4a8 | Context: 131072 | Chip: M50 | Build: 1.4.0 | Status: Temporary

Google
Gemma4_26b-a4b_128k_b1_2core_1chip(768*2520)
26 gguf w4a8 M50 131,072 视觉
文本 文本
HLIECpp HLIEPython HLIELLama

Type: Gemma4 | Size: 26 | Quant: w4a8 | Context: 131072 | Chip: M50 | Build: 1.4.0 | Status: Stable

Google
Gemma4_26b-a4b_256k_b1_2core_1chip(768*2520)
26 gguf w4a8 M50 262,144 视觉
文本 文本
HLIECpp HLIEPython HLIELLama

Type: Gemma4 | Size: 26 | Quant: w4a8 | Context: 262144 | Chip: M50 | Build: 1.4.0 | Status: Stable

Google
Gemma4_26b-a4b_256k_b1_2core_2chips(768*2520)
26 gguf w4a8 M50 262,144 视觉
文本 文本
HLIECpp HLIEPython HLIELLama

Type: Gemma4 | Size: 26 | Quant: w4a8 | Context: 262144 | Chip: M50 | Build: 1.4.0 | Status: Stable

Google
Gemma4_31b_256k_b1_2core_2chips(768*2520)
31 gguf w4a8 M50 262,144 视觉
文本 文本
HLIECpp HLIEPython HLIELLama

Type: Gemma4 | Size: 31 | Quant: w4a8 | Context: 262144 | Chip: M50 | Build: 1.4.0 | Status: Stable

Google
Gemma4_31b_256k_b1_2core_4chips(768*2520)
31 gguf w4a8 M50 262,144 视觉
文本 文本
HLIECpp HLIEPython HLIELLama

Type: Gemma4 | Size: 31 | Quant: w4a8 | Context: 262144 | Chip: M50 | Build: 1.4.0 | Status: Temporary

Google
Gemma4_4b_128k_b1_2core_1chip()
4 gguf w4a8 M50 131,072 视觉
文本 文本
HLIECpp HLIEPython HLIELLama

Type: Gemma4 | Size: 4 | Quant: w4a8 | Context: 131072 | Chip: M50 | Build: 1.4.0 | Status: Stable

智谱
GLM-ASR_1.9b__b1_2core_1chip()
1.9 gguf w4a8 M50 视觉
音频 文本
HLIECpp HLIEPython HLIELLama

Type: GLM-ASR | Size: 1.9 | Quant: w4a8 | Chip: M50 | Build: 1.4.0 | Status: Stable

智谱
GLM-OCR_0.9b_8k_b1_2core_1chip()
0.9 gguf w4a8 M50 8,192 视觉
图像 文本
HLIECpp HLIEPython HLIELLama

Type: GLM-OCR | Size: 0.9 | Quant: w4a8 | Context: 8192 | Chip: M50 | Build: 1.4.0 | Status: Stable

智谱
GLM-OCR_0.9b_8k_b1_2core_1chip(672*672)
0.9 gguf w4a8 M50 8,192 视觉
图像 文本
HLIECpp HLIEPython HLIELLama

Type: GLM-OCR | Size: 0.9 | Quant: w4a8 | Context: 8192 | Chip: M50 | Build: 1.3.0 | Status: Stable

OpenAI
GPT-OSS_120b_128k_b1_2core_2chips
120 gguf w4a8 M50 131,072
文本 文本
HLIECpp HLIEPython HLIELLama

Type: GPT-OSS | Size: 120 | Quant: w4a8 | Context: 131072 | Chip: M50 | Build: 1.4.0 | Status: Temporary

OpenAI
GPT-OSS_20b_256k_b1_2core_1chip
20 gguf w4a8 M50 262,144
文本 文本
HLIECpp HLIEPython HLIELLama

Type: GPT-OSS | Size: 20 | Quant: w4a8 | Context: 262144 | Chip: M50 | Build: 1.4.0 | Status: Stable

Alibaba
Gte_1.5b_2k_b1_2core_1chip
1.5 gguf w4a8 M50 2,048
文本 向量
HLIECpp HLIEPython HLIELLama

Type: Gte | Size: 1.5 | Quant: w4a8 | Context: 2048 | Chip: M50 | Build: 1.4.0 | Status: Stable

MINICPM
MiniCPM_1.3B_256k_b1_2core_1chip(dynamic)
1.3 gguf w8a8 M50 262,144 视觉
文本 文本

Type: MiniCPM | Size: 1.3 | Quant: w8a8 | Context: 262144 | Chip: M50 | Build: 1.5.0 | Status: Stable

QWEN-AGENTWORLD
Qwen-AgentWorld_35b-a3b_256k_b1_2core_1chip
35 gguf w4a8 M50 262,144
文本 文本
HLIECpp HLIEPython HLIELLama

Type: Qwen-AgentWorld | Size: 35 | Quant: w4a8 | Context: 262144 | Chip: M50 | Build: 1.4.0 | Status: Stable

QWEN-AGENTWORLD
Qwen-AgentWorld_35b-a3b_32k_b1_2core_1chip
35 gguf w4a8 M50 32,768
文本 文本
HLIECpp HLIEPython HLIELLama

Type: Qwen-AgentWorld | Size: 35 | Quant: w4a8 | Context: 32768 | Chip: M50 | Status: Temporary