Palabra × OpenAI Agents SDK
Palabra Realtime STT and TTS as models for the VoicePipeline
of the OpenAI Agents SDK. Speech recognition and synthesis go
through Palabra instead of gpt-4o-transcribe / tts-1; the agent (LLM, tools, handoffs) stays as it is.
pip install palabra-ai-openai-agents
export PALABRA_API_KEY=plbr_... # https://platform.palabra.ai/api-keys
Usage
from agents import Agent
from agents.voice import SingleAgentVoiceWorkflow, VoicePipeline
from palabra_openai_agents import PalabraSTTModel, PalabraTTSModel
agent = Agent(name="Assistant", instructions="You are a helpful voice assistant.")
pipeline = VoicePipeline(
workflow=SingleAgentVoiceWorkflow(agent),
stt_model=PalabraSTTModel(language="en"), # None = server-side language detection
tts_model=PalabraTTSModel(language="en", voice_id="default_low"),
)
Then use the pipeline as usual: await pipeline.run(AudioInput(buffer=..., frame_rate=...)) for a
recorded turn, or StreamedAudioInput for a live microphone. Or hand out both models through the config:
from agents.voice import VoicePipelineConfig
from palabra_openai_agents import PalabraVoiceModelProvider
config = VoicePipelineConfig(model_provider=PalabraVoiceModelProvider(stt_language="en", tts_language="en"))
pipeline = VoicePipeline(workflow=..., config=config)
Examples: examples/voice_agent_mic.py (microphone → agent → speakers,
needs sounddevice and OPENAI_API_KEY) and examples/voice_agent_file.py
(WAV question in, WAV answer out).
How it maps
| SDK interface | Implementation |
|---|---|
STTModel.transcribe(AudioInput) |
the buffer (int16 or float32, any rate) is streamed at its native frame_rate, paced to real time; then finalize is sent and the finals up to the <fin> marker are joined. Takes about as long as the audio. |
STTModel.create_session(StreamedAudioInput) |
a pump task feeds queued chunks (24 kHz by SDK convention, input_frame_rate= to change) into one STT WebSocket; transcribe_turns() yields every final transcript as a turn (turn_gap_s= merges finals that arrive close together). add_audio(None) ends the input with finalize. |
TTSModel.run(text) |
one TTS WebSocket per call, pcm s16le mono at 24 kHz — the format VoicePipeline expects. Text is split into sentences under the 1024-character limit and the audio streamed in order. TTSModelSettings.speed is honoured (clamped to 0.1–2.0); settings.voice (OpenAI names) is ignored — pick a Palabra voice with voice_id=. |
Options of PalabraSTTModel: language, input_frame_rate, turn_gap_s, finalize_timeout_s,
api_key / client (a configured palabra_ai.Palabra, e.g. for region="us").
Tests
pip install -e ".[dev]"
pytest -m "not live" # offline, scripted fake of the Palabra client
PALABRA_API_KEY=plbr_... pytest -m live -s # against the API; speech is produced by Palabra TTS
PALABRA_TEST_WAV=my.wav PALABRA_TEST_LANGUAGE=ru pytest -m live -k user_recording -s
Notes
- One API Key is not limited to a single live STT session: parallel
create_session/transcribecalls each open their own session with the same key.
Metadata
Release files for palabra-ai-openai-agents 0.1.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| palabra_ai_openai_agents-0.1.0.tar.gz | 14.9 kB | Without provenance | Details for palabra_ai_openai_agents-0.1.0.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| palabra_ai_openai_agents-0.1.0-py3-none-any.whl | Python 3 | none | any | Details for palabra_ai_openai_agents-0.1.0-py3-none-any.whl |
Total release size: 27.6 kB
Release files / palabra_ai_openai_agents-0.1.0.tar.gz
| Download URL | palabra_ai_openai_agents-0.1.0.tar.gz |
|---|---|
| Size | 14.9 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
372712d3bc7da2f46f7765fa3f33f1a3a0f20984ee540b6bd55ae051288636b4
|
|
BLAKE2b-256 checksum How to use checksums |
9211a8293590cf3a61686da8c23efe540270df0e22374850f9668c767ee89bf3
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/7.0.0 CPython/3.13.12
|
Release files / palabra_ai_openai_agents-0.1.0-py3-none-any.whl
| Download URL | palabra_ai_openai_agents-0.1.0-py3-none-any.whl |
|---|---|
| Size | 12.8 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
521898b61469057600d0265c2764f4cd75ea1d32bc9898ae2c6ebc210a51476b
|
|
BLAKE2b-256 checksum How to use checksums |
352adddd16ac0fc5a915262376fb91e9d840f6ec6fe93c88bfba0f67aa612b5d
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/7.0.0 CPython/3.13.12
|
Security for palabra-ai-openai-agents 0.1.0
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.