GOESB — Global Open Edge Speech Benchmark · pronounced “goose-bee”

Leaderboard

Objective, reproducible benchmarks for local (edge) speech AI. Every result here re-verified its hash and signature before being accepted — see why you can trust the numbers. New here? Read the docs.

This is every verified result, filterable and sortable. Looking for a single best recommendation for your constraints instead? Try the homepage →

Lowest energy Fastest (lowest RTF)Most accurate (lowest WER)

StreamingStreaming results only — real-time factor doesn't apply, so batch results never appear here.

Any Batch Streaming Concurrency

Sorted by most recently submitted · 9 results · click a metric below to sort by it

Profile	Runtime / Model	Hardware	Pack	Metrics	Submitted	Details
whispercpp-large-v3-de-streaming	whisper-cpp / whisper-large-v3metal	apple-m1-pro	fleurs-de	wer: 0.0720 ratio ram_mb: 4074.72 MB cpu_pct: 8.89 %real_time_factor: 1.06 ratio update_frequency: 0.9578 Hz partial_stability: 0.8632 ratio first_final_latency: 5443.45 ms end_of_speech_latency: 2701.40 ms first_partial_latency: 1713.80 ms streaming_responsiveness: 0.4530 index	2026-08-03T14:15:52Z	details Profile whispercpp-large-v3-de-streaming@1.0.0 Pack fleurs-de@1.0.0 Repeats 2 Backend metal chunk_ms 1000 threads 4
whispercpp-large-v3-es-streaming	whisper-cpp / whisper-large-v3metal	apple-m1-pro	fleurs-es	wer: 0.0399 ratio ram_mb: 4063.33 MB cpu_pct: 8.20 %real_time_factor: 1.07 ratio update_frequency: 0.9534 Hz partial_stability: 0.9450 ratio first_final_latency: 5233.96 ms end_of_speech_latency: 2160.05 ms first_partial_latency: 1734.84 ms streaming_responsiveness: 0.5007 index	2026-08-03T14:27:31Z	details Profile whispercpp-large-v3-es-streaming@1.0.0 Pack fleurs-es@1.0.0 Repeats 2 Backend metal chunk_ms 1000 threads 4
whispercpp-large-v3-fr-streaming	whisper-cpp / whisper-large-v3metal	apple-m1-pro	fleurs-fr	wer: 0.0704 ratio ram_mb: 4064.90 MB cpu_pct: 9.44 %real_time_factor: 1.12 ratio update_frequency: 1.01 Hz partial_stability: 0.8612 ratio first_final_latency: 4992.39 ms end_of_speech_latency: 1877.41 ms first_partial_latency: 1703.40 ms streaming_responsiveness: 0.4817 index	2026-08-03T14:33:07Z	details Profile whispercpp-large-v3-fr-streaming@1.0.0 Pack fleurs-fr@1.0.0 Repeats 2 Backend metal chunk_ms 1000 threads 4
whispercpp-large-v3-nl-streaming	whisper-cpp / whisper-large-v3metal	apple-m1-pro	fleurs-nl	wer: 0.0985 ratio ram_mb: 4065.51 MB cpu_pct: 9.66 %real_time_factor: 1.13 ratio update_frequency: 1.05 Hz partial_stability: 0.8591 ratio first_final_latency: 4726.40 ms end_of_speech_latency: 1804.48 ms first_partial_latency: 1678.25 ms streaming_responsiveness: 0.5207 index	2026-08-03T14:37:46Z	details Profile whispercpp-large-v3-nl-streaming@1.0.0 Pack fleurs-nl@1.0.0 Repeats 2 Backend metal chunk_ms 1000 threads 4
whispercpp-large-v3-pt-streaming	whisper-cpp / whisper-large-v3metal	apple-m1-pro	fleurs-pt	wer: 0.0792 ratio ram_mb: 4009.50 MB cpu_pct: 9.26 %real_time_factor: 1.11 ratio update_frequency: 1.04 Hz partial_stability: 0.8740 ratio first_final_latency: 4257.48 ms end_of_speech_latency: 2071.59 ms first_partial_latency: 1602.57 ms streaming_responsiveness: 0.5565 index	2026-08-03T14:44:58Z	details Profile whispercpp-large-v3-pt-streaming@1.0.0 Pack fleurs-pt@1.0.0 Repeats 2 Backend metal chunk_ms 1000 threads 4
whispercpp-medium-de-streaming	whisper-cpp / whisper-mediummetal	apple-m1-pro	fleurs-de	wer: 0.0640 ratio ram_mb: 2228.53 MB cpu_pct: 14.81 %real_time_factor: 0.7309 ratio update_frequency: 1.01 Hz partial_stability: 0.8466 ratio first_final_latency: 5170.98 ms end_of_speech_latency: 2339.54 ms first_partial_latency: 2049.79 ms streaming_responsiveness: 0.5012 index	2026-08-03T14:49:06Z	details Profile whispercpp-medium-de-streaming@1.0.0 Pack fleurs-de@1.0.0 Repeats 2 Backend metal chunk_ms 1000 threads 4
whispercpp-medium-es-streaming	whisper-cpp / whisper-mediummetal	apple-m1-pro	fleurs-es	wer: 0.0570 ratio ram_mb: 2227.63 MB cpu_pct: 14.95 %real_time_factor: 0.7302 ratio update_frequency: 0.9686 Hz partial_stability: 0.8806 ratio first_final_latency: 4882.64 ms end_of_speech_latency: 1361.95 ms first_partial_latency: 1373.69 ms streaming_responsiveness: 0.5930 index	2026-08-03T14:56:23Z	details Profile whispercpp-medium-es-streaming@1.0.0 Pack fleurs-es@1.0.0 Repeats 2 Backend metal chunk_ms 1000 threads 4
whispercpp-medium-pt-streaming	whisper-cpp / whisper-mediummetal	apple-m1-pro	fleurs-pt	wer: 0.0574 ratio ram_mb: 2228.35 MB cpu_pct: 12.49 %real_time_factor: 0.6280 ratio update_frequency: 0.9966 Hz partial_stability: 0.8776 ratio first_final_latency: 4379.89 ms end_of_speech_latency: 1622.26 ms first_partial_latency: 1644.79 ms streaming_responsiveness: 0.5670 index	2026-08-03T15:07:22Z	details Profile whispercpp-medium-pt-streaming@1.0.0 Pack fleurs-pt@1.0.0 Repeats 2 Backend metal chunk_ms 1000 threads 4
whispercpp-small-es-streaming	whisper-cpp / whisper-smallmetal	apple-m1-pro	fleurs-es	wer: 0.0883 ratio ram_mb: 888.32 MB cpu_pct: 19.66 %real_time_factor: 0.2380 ratio update_frequency: 1.01 Hz partial_stability: 0.9030 ratio first_final_latency: 4564.12 ms end_of_speech_latency: 1407.30 ms first_partial_latency: 1165.19 ms streaming_responsiveness: 0.7540 index	2026-08-03T15:11:17Z	details Profile whispercpp-small-es-streaming@1.0.0 Pack fleurs-es@1.0.0 Repeats 2 Backend metal chunk_ms 1000 threads 4

Constant across this view: chunk_ms 1000 · threads 4

* profile default