IBM: Granite 4.1 8B (~519ms)
Google: Gemma 3 12B (~407ms)
DeepSeek: DeepSeek V4 Flash (~467ms)
IBM: Granite 4.0 Micro (~432ms)
Google: Gemma 3 4B (~606ms)
Mistral: Ministral 3 8B 2512 (~384ms)
Mistral: Mistral Small 3 (~208ms)
Amazon: Nova Micro 1.0 (~549ms)
Microsoft: Phi 4 (~207ms)
Amazon: Nova Lite 1.0 (~305ms)
Cohere: Command R7B (12-2024) (~757ms)
Google: Gemma 3 27B (~996ms)
inclusionAI: Ling-2.6-flash (~599ms)
Meta: Llama 3.3 70B Instruct (~141ms)
Mistral: Ministral 3 3B 2512 (~353ms)
Nex AGI: Nex-N2-Mini (~670ms)
Mistral: Mistral Small 3.2 24B (~464ms)
OpenAI: gpt-oss-20b (~723ms)
NVIDIA: Nemotron 3 Nano 30B A3B (~162ms)
Qwen: Qwen3 14B (~241ms)