positron-ai/DeepSeek-V4-Flash-0731
304B • Updated
AI & ML interests
Hardware and software platforms for faster inference.
Recent Activity
View all activity
Organization Card
Positron AI is building novel systems and software to accelerate LLM inference. Faster inference at useful batch sizes on a fraction the power of the competition.
models 28
positron-ai/Kimi-Linear-48B-A3B-Instruct
49B • Updated
positron-ai/GLM-4.7-Flash
31B • Updated
positron-ai/Qwen3-0.6B
0.8B • Updated
positron-ai/Qwen3.5-4B
5B • Updated
positron-ai/meta-llama_Llama-3.2-1B-Instruct-ingest-best-gptq
Text Generation • 1B • Updated • 16
positron-ai/openai_gpt-oss-20b-hf-bf16
Text Generation • 21B • Updated • 568
positron-ai/google_gemma-4-31B-it-ingest-best-gptq-permuted
Text Generation • 31B • Updated • 127
positron-ai/meta-llama_Llama-3.2-1B-Instruct-tron-best-gptq-permuted
Text Generation • 1B • Updated • 15
positron-ai/meta-llama_Llama-3.1-8B-Instruct-ingest-best-gptq-permuted
Text Generation • 8B • Updated • 19
datasets 0
None public yet