FrontierRL
Frontier Inference
Decoding throughput, tokens per second.
FrontierRL Star
0t/s
GPT-5.6 on Cerebras
0t/s
Xiaomi
0t/s
DeepSeek
0t/s
Claude Opus 5 Thinking
0t/s
00
4X faster than the frontier lead
Star outputs measured on workstation GPUs, not datacenter-grade.
At Budget-Level Prices
Calculated per 1M output tokens, at each model’s published price.
How far does $100 go?
FrontierRL Star $0.40 / M
0M
Gemini 3.1 Pro $12 / M
0M
Claude Sonnet 5 $15 / M
0M
Claude Opus 5 $25 / M
0M
GPT-5.6 Sol $30 / M
0M
00
~75× more tokens per dollar
Watch Star run.
Why Star is different
Not another LLM.
A new architecture, built for agents from day one.
It understands live video, audio, and text, then acts across real software—fast enough for the moment, efficient enough to run continuously.
Programs
MIT TNT Research 
Together.ai Accelerator