AI models.
Optimized for agentic workflows by Ajelix
GLM-5.3-Flash
Z.ai
ModeAjelix Optimized
Region
EU (LV)
EU (LV)Parameters320B total · 18B active
Input tokens, 1M€0.30
Output tokens, 1M€0.90
Cache read€0.035
Tokens per secUp to 170
Context1M
DeepSeek-V4-Flash
DeepSeek
ModeAjelix Optimized
Region EU (LV)
EU (LV)Parameters284B total · 13B active
Input tokens, 1M€0.25
Output tokens, 1M€0.75
Cache read€0.035
Tokens per secUp to 200
Context1M
Qwen3.8-27B
Qwen
ModeAjelix Optimized
Region EU (LV)
EU (LV)Parameters27B (dense)
Input tokens, 1M€0.50
Output tokens, 1M€3.00
Cache read€0.05
Tokens per secUp to 120
Context262K
Qwen3-Embedding-4B
Qwen
ModeAjelix Optimized
Region EU (LV)
EU (LV)Parameters4B (dense)
Tokens, 1M€0.05
TypeEmbedding