Together AI / Llama 3.3 70B on MMLU

Together AI hosting Llama 3.3 70B on MMLU. Completed at 4/15/2026, 10:02:20 PM. 85.0%. 885 ms p50 latency.

Run Summary

This run is completed. Benchscope recorded 100 samples with 100 completed samples. Prompt mode: canonical. Sample scope: random.

Benchmark and Endpoint

Model family: Llama 3.3 70B. Provider: Together AI. Benchmark: MMLU. Compare this run with other public runs on the same benchmark and model pages.

Benchscope is a JavaScript app. If the interactive interface does not load, enable JavaScript or use the links above for the main public sections.