Occamy
by Accio Lab · Model · Language & code models
Occamy 1.0 is a 35B-A3B agentic post-train of Qwen with open weights and training recipe; smallest GGUF about 12 GB
AccessOpen weightsNeeds~12 GBDownload12 GBLicenseapache-2.0
Evidence
- Sep 20Best average among similar-size models on agentic benchmarks; competitive with some much larger models