Software
or, TINA. Neural archiphonemic transcriber for Brazilian Portuguese, under continuous development.
No uploaded files. Single-word transcription requests are transient.
Enter a word above and press Transcribe.
These signals reflect how stably and confidently TINA decoded this word. They are not a phonotactic legality score.
TINA reads an orthographic word in Brazilian Portuguese and produces a structured phonetic transcription. The model determines syllable structure and stress first, fills in phoneme content, then applies phonological rules that generate dialect-sensitive variants.
The output uses a partially abstract notation where certain symbols represent classes of sounds whose realization depends on dialect and context. TINA generalizes to unseen words but systematically avoids antepenultimate stress. Outputs should always be verified manually.
This tool is still under active development and must be used cautiously.
Use the registered TINA software reference when citing this web app.
Barroso, A. M. (2026). TINA (Transcrição Inteligente, Notação Automática) (Registration No. BR5120260028949) [Computer software]. Instituto Nacional da Propriedade Industrial.
BARROSO, A. M. TINA (Transcrição Inteligente, Notação Automática). 2026. Patente: Programa de computador. Registro n. BR5120260028949, registrado em 09 abr. 2026. Instituto Nacional da Propriedade Industrial (INPI).