Tasks and inputs
Gemini 3.8 Live is listed for Audio-to-audio. It accepts text, image, audio, video and produces audio, text. Use the publisher’s input format and preprocessing for your chosen deployment.
Gemini 3.8 Live is Google's model for interactive audio conversations with text, image, audio and video input. Its Live API response modality is audio; text is available through transcription.
The specification score summarizes recorded features and limits. It is not a benchmark result.
Recorded modality support shows what this model can accept and produce. An undocumented modality is shown as unknown rather than assumed to be unsupported.
Accepts
Produces
Gemini 3.8 Live is Google's model for interactive audio conversations with text, image, audio and video input. Its Live API response modality is audio; text is available through transcription.
Live is relevant when an application must listen and respond during an ongoing session. It uses a bidirectional connection and supports asynchronous function calls. This is a different integration from sending a prompt to Gemini Flash and waiting for a text response.
The documented input allowance is 131,072 tokens and the output allowance is 65,536 tokens. It does not provide structured JSON output, code execution or ordinary prompt caching. Do not substitute it for a text model merely because the version numbers match.
Measure response onset, interruptions, recovery after a connection failure and whether the agent resumes the right task after a tool result arrives. Include noisy microphones, accents and users who correct themselves halfway through a request. Check both audio and transcript handling in the finished interface.
The pricing schedule separates audio and text tokens. A single input/output text rate would misrepresent a voice session, so this profile leaves that combined price unset. Inspect the current source pricing before estimating a session budget. Compare Google's models or browse the audio collection.
Compare the model on the work you need to complete.
Gemini 3.8 Live is listed for Audio-to-audio. It accepts text, image, audio, video and produces audio, text. Use the publisher’s input format and preprocessing for your chosen deployment.
Documented limits: 131.07K context and 65.54K maximum output. Compare pronunciation, voice consistency and time to the first playable audio. Check the service’s billing unit for the actual script.
Use the model identifier and account requirements for the selected API route. Confirm its request limits and output format with one complete task before connecting a larger workflow.
These records share a directory category. Check the source card, output type and deployment requirements before treating a model as a replacement.
Alibaba
Qwen-Audio-3.1-Realtime-Plus supports duplex speech conversation, voice cloning, tools and a 262K-token context through its realtime interface.
Aratako
Aratako/MioCodec-25Hz-24kHz is a source-linked Hugging Face repository assigned to Audio-to-audio. Its recorded task interface accepts audio and produces audio.
Aratako
Aratako/MioCodec-25Hz-44.1kHz-v2 is a source-linked Hugging Face repository assigned to Audio-to-audio. Its recorded task interface accepts audio and produces audio.
aufklarer
aufklarer/DeepFilterNet3-CoreML is a source-linked Hugging Face repository assigned to Audio-to-audio. Its recorded task interface accepts audio and produces audio.
aufklarer
aufklarer/PersonaPlex-7B-MLX-4bit is a source-linked Hugging Face repository assigned to Audio-to-audio. Its recorded task interface accepts audio and produces audio.
aufklarer
aufklarer/Sidon-CoreML is a source-linked Hugging Face repository assigned to Audio-to-audio. Its recorded task interface accepts audio and produces audio.