Aguasto / models
Models. In context.
A curated open-weight index connecting checkpoint identity, licensing, serving software and memory requirements. These are research references, not available Aguasto model endpoints.
7 of 7 references
GLM-4.5
Mixture of experts; hybrid thinking and non-thinking text model. MIT (model weights). Public checkpoint reference.
Qwen3-235B-A22B
Causal mixture-of-experts language model; thinking and non-thinking modes. Apache 2.0. Public checkpoint reference.
DeepSeek-R1
Mixture-of-experts reasoning language model based on DeepSeek-V3-Base. MIT. Public checkpoint reference.
Llama 3.3 70B Instruct
Dense autoregressive transformer with grouped-query attention. Llama 3.3 Community License. Publisher access gate applies.
Gemma 3 27B IT
Multimodal transformer; text decoder with vision encoder. Gemma Terms of Use. Publisher access gate applies.
gpt-oss-120b
Mixture-of-experts reasoning language model. Apache 2.0. Public checkpoint reference.
Kimi-K2-Instruct
Mixture-of-experts language model with Multi-head Latent Attention. Modified MIT. Public checkpoint reference.