A systematic evaluation of audio multimodal model capabilities using structured test prompts, a natural voice recording, and programmatic output generation.
| Parameter | Value |
|---|---|
| Model under test | gemini-3.1-flash-lite-preview |
| Audio sample | FLAC mono 24kHz, 20m 54s |
| Transcript baseline | AssemblyAI (97.4% confidence) |
| Prompts executed | 49 across 13 categories |
| Date | 26 March 2026 |
| Author | Daniel Rosehill |
| Assisted by | Claude (Opus 4.6) |
The audio sample is an informal, unscripted 21-minute voice recording by a male Irish-accented speaker in his late 30s, recorded on a smartphone in a residential room in Jerusalem. Topics range from personal experience during the Iran–Israel conflict to AI podcast production, voice cloning experiments, and EQ analysis.
Model evaluated: Google Gemini 3.1 Flash Lite (Preview)
- Speaker identification & accent analysis — Correctly identified Irish male, late 30s, with detailed phonetic observations across multiple prompts
- Content comprehension — Accurately extracted topics, named entities, geographic references, and technical terminology from unscripted speech
- Emotional tone & deception analysis — Appropriately characterised the speech as casual and non-deceptive, noting fatigue consistent with the speaker's own description
- Audio engineering — Technically sound EQ recommendations, plausible microphone type and room acoustics inferences
- Ethical responsibility — Consistently included appropriate disclaimers on sensitive prompts (mental health, drug influence, deception)
- Quantitative metrics — WPM analysis produced structured output but interval breakdowns appear estimated rather than measured
- Speaker height estimation — Reasonable estimate with good caveats, but cited acoustic evidence reads as generic rather than specific to this recording
- Adversarial prompts — True age detection relied on the speaker's own statements rather than independent vocal analysis
- Acoustic signal analysis — Tasks requiring spectral or temporal measurement (valence-arousal mapping, emotional peaks with timestamps) produced plausible narratives without demonstrable acoustic grounding
- Weather inference & voice matching — Minimal outputs that did not attempt conditional reasoning from available cues
Gemini 3.1 Flash Lite is strong at content-driven audio understanding (who is speaking, what they said, the emotional register, the environment) but limited in signal-level acoustic analysis (precise measurements, spectral characteristics). Quantitative outputs should be treated as informed estimates, not measurements. For a lightweight model variant, this is an impressive baseline — all 49 prompts completed without failures, with coherent and relevant outputs throughout.
2026-03-26/26_03_2026_16_08.flac— Full recording (~30MB, mono 24kHz FLAC, 20m 54s)2026-03-26/short-sample.flac— Shorter excerpttranscript.md— Timestamped transcript (AssemblyAI, 97.4% confidence)
49 implemented prompts across 13 categories:
| Category | Count | Examples |
|---|---|---|
| Speaker Analysis | 8 | accent, phonetic-analysis, speech-patterns, hybrid-accent-analysis |
| Audio Engineering | 6 | eq-recommendation, microphone-type-guess, room-acoustics-estimation |
| Emotion & Sentiment | 5 | emotional-tone, valence-arousal-mapping, tone-of-voice-description |
| Speaker Demographics | 4 | gender-determination, true-age-detection, education-level-estimation |
| Health & Wellness | 4 | inebriation-detection, drug-influence-detection, hydration-assessment |
| Environment | 4 | indoor-outdoor-classification, background-noise-classification |
| Speech Metrics | 3 | wpm-analysis, dictation-coaching, speech-recognition-model-ranking |
| Forensic Audio | 3 | deception-detection, deepfake-detection, insincerity-timestamps |
| Voice Cloning | 2 | tts-cloning-notes, clonability-assessment |
| Content Analysis | 2 | words-vs-tone-deviation, address-pattern-rating |
| Language Learning | 2 | hebrew-phonetic-difficulty, easiest-foreign-language |
| Production | 1 | voiceover-potential |
| Speaker ID | 1 | celebrity-voice-match |
Full index: test-prompts/prompt-index.json
Raw model outputs (markdown) for all 49 prompts, generated by run-prompts.py.
report.pdf— Compiled PDF with cover page, experiment introduction, findings summary, table of contents, and all 49 prompt–output pairsgenerate-pdf.py— Script to regenerate the PDF from current data
# Run all prompts against the model
python run-prompts.py
# Run a specific prompt
python run-prompts.py --only accent
# Regenerate the PDF report
.venv/bin/python generate-pdf.pyTest prompts and outputs are provided for research and reference purposes.