Swahili Evaluation of Audio AI Models Swahili Evaluation of Audio AI Models By Gooey.AI, CLEAR Global and the Gates Foundation • 30 Aug 2026 • Gooey.AI/language 0 2 4 6 8 10 12 0.6 0.7 0.8 0.9 1.0 Accuracy Latency Lower is better More accurate Gemini 3.1 Pro — Accuracy 0.95, Latency 5.79Gemini 3.1 ProFastestMost Accurate Gemini 3.1 Pro + Omni — Accuracy 0.92, Latency 9.33Gemini 3.1 Pro + Omni Gemini 3.1 Pro + Intron — Accuracy 0.91, Latency 10.73Gemini 3.1 Pro + Intron Gemini 3.1 Pro + GPT Transcribe — Accuracy 0.91, Latency 8.23Gemini 3.1 Pro + GPT Transcribe GPT 5.6 Sol + Omni — Accuracy 0.9, Latency 7GPT 5.6 Sol + Omni GPT 5.6 Sol + Intron — Accuracy 0.88, Latency 9.89GPT 5.6 Sol + Intron GPT 5.6 Sol + GPT Transcribe — Accuracy 0.87, Latency 6.39GPT 5.6 Sol + GPT Transcribe Claude Sonnet 5 + Omni — Accuracy 0.91, Latency 6.72Claude Sonnet 5 + Omni Claude Sonnet 5 + Intron — Accuracy 0.89, Latency 8.36Claude Sonnet 5 + Intron Claude Sonnet 5 + GPT Transcribe — Accuracy 0.84, Latency 5.85Claude Sonnet 5 + GPT Transcribe Kimi K3 + Omni — Accuracy 0.68, Latency 7.12Kimi K3 + Omni Kimi K3 + Intron — Accuracy 0.65, Latency 9.29Kimi K3 + Intron Kimi K3 + GPT Transcribe — Accuracy 0.7, Latency 6.71Kimi K3 + GPT Transcribe DeepSeek V4 + Omni — Accuracy 0.93, Latency 7.55DeepSeek V4 + Omni DeepSeek V4 + Intron — Accuracy 0.93, Latency 9.7DeepSeek V4 + Intron DeepSeek V4 + GPT Transcribe — Accuracy 0.87, Latency 5.89DeepSeek V4 + GPT Transcribe Gemini 3.7 Flash + Omni — Accuracy 0.84, Latency 7.61Gemini 3.7 Flash + Omni Gemini 3.7 Flash + Intron — Accuracy 0.85, Latency 10.02Gemini 3.7 Flash + Intron Gemini 3.7 Flash + GPT Transcribe — Accuracy 0.88, Latency 7.08Gemini 3.7 Flash + GPT Transcribe Gemini 3.1 Pro — Accuracy 0.95, Latency 5.790 Gemini 3.1 Pro + Omni — Accuracy 0.92, Latency 9.331 Gemini 3.1 Pro + Intron — Accuracy 0.91, Latency 10.732 Gemini 3.1 Pro + GPT Transcribe — Accuracy 0.91, Latency 8.233 GPT 5.6 Sol + Omni — Accuracy 0.9, Latency 74 GPT 5.6 Sol + Intron — Accuracy 0.88, Latency 9.895 GPT 5.6 Sol + GPT Transcribe — Accuracy 0.87, Latency 6.396 Claude Sonnet 5 + Omni — Accuracy 0.91, Latency 6.727 Claude Sonnet 5 + Intron — Accuracy 0.89, Latency 8.368 Claude Sonnet 5 + GPT Transcribe — Accuracy 0.84, Latency 5.859 Kimi K3 + Omni — Accuracy 0.68, Latency 7.1210 Kimi K3 + Intron — Accuracy 0.65, Latency 9.2911 Kimi K3 + GPT Transcribe — Accuracy 0.7, Latency 6.7112 DeepSeek V4 + Omni — Accuracy 0.93, Latency 7.5513 DeepSeek V4 + Intron — Accuracy 0.93, Latency 9.714 DeepSeek V4 + GPT Transcribe — Accuracy 0.87, Latency 5.8915 Gemini 3.7 Flash + Omni — Accuracy 0.84, Latency 7.6116 Gemini 3.7 Flash + Intron — Accuracy 0.85, Latency 10.0217 Gemini 3.7 Flash + GPT Transcribe — Accuracy 0.88, Latency 7.0818 © 2026 Gooey.AI