by ByteDancePremium
ByteDance flagship — 76.5% SWE-Bench, 98.3% AIME 2025, hour-long video understanding
Detailed content for this model is coming soon.
MiniMax
Latest M-series — 1M context, agentic coding, tool use, vision + video input, interleaved thinking
Anthropic
Most capable widely released Claude — currently unavailable due to US export controls. Use Claude Opus 4.8 instead.
Anthropic
Hybrid reasoning model with 1M context, top-tier coding and agentic performance
Read full article