We’re excited to announce that we’ve expanded our supported AI model library with three new additions for enhanced versatility and performance.
New Models Added
Qwen2.5 1.5B Instruct
Available in multiple quantization formats (Q2_K → FP16) for diverse performance/memory trade-offs.
Llama 3.2 3B Instruct
Includes IQ, Q3, Q4, Q5, Q6, Q8, and F16 variants for flexible deployment.
Tesslate Tessa T1 3B
Wide range of quantization options from IQ2 to BF16 for optimal inference performance.