StepFun's Step 3.7 Flash model leads AI output speed benchmarks

StepFun, an AI company, has released its latest open-source base model, Step 3.7 Flash. This model has achieved a leading output speed of 409 tokens/s on the Artificial Analysis platform's Output Speed leaderboard. It also excels in other key metrics such as end-to-end response time, intelligence versus output speed, and speed-to-price ratio. AI

IMPACT Sets a new benchmark for AI model output speed, potentially influencing future model development and deployment strategies.

RANK_REASON The cluster reports on a new open-source base model release and its benchmark performance. [lever_c_demoted from research: ic=1 ai=1.0]

Read on 36氪 (36Kr) →

• Artificial Analysis
• Step 3.7 Flash
• StepFun

• model release
• product

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →