awesome_benchmarks.md

Long-context benchmarks

• Mon, 7 Jul 2025 LOOM-Scope: a comprehensive and efficient LOng-cOntext Model evaluation framework
  • Our platform supports 22 long-context benchmarks and more than 140 long-context tasks.
  • https://loomscope.github.io/ 20250711 1. Qwen3-14B 2. Qwen3-30B-A3B 3. Meta-Llama-3.1-8B-Instruct 4. c4ai-command-r7b-12-2024 4. glm-4-9b-chat 6. Qwen3-8B