transcription
Ultravox v0.6 70B
Ultravox v0.6 70BUltravox is a multimodal model that can consume both speech and text as input, generating output text as usual. This uses Llama 3.3 70B Instruct as its backbone.
Model details
View repositoryExample usage
Since this is a multimodal model, it accepts text and/or audio. The audio is downloaded from a public source.
The output JSON object contains a key called content which represents the usual output text.
Input
1from openai import OpenAI
2
3model_id = "jwdp26kw" # Replace with your model ID from Baseten's model dashboard
4
5client = OpenAI(
6 api_key="YOUR-API-KEY",
7 base_url=f"https://model-{model_id}.api.baseten.co/environments/production/sync/v1"
8)
9
10response = client.chat.completions.create(
11 model="",
12 messages=[
13 {
14 "role": "user",
15 "content": [
16 {
17 "type": "text",
18 "text": "What is Lydia like?"
19 },
20 {
21 "type": "audio_url",
22 "audio_url": {"url": "https://baseten-public.s3.us-west-2.amazonaws.com/fred-audio-tests/real.mp3"}
23 }
24 ]
25 }
26 ]
27)
28
29print(response)JSON output
1{
2 "id": "143",
3 "choices": [
4 {
5 "finish_reason": "stop",
6 "index": 0,
7 "logprobs": null,
8 "message": {
9 "content": "[Model output here]",
10 "role": "assistant",
11 "audio": null,
12 "function_call": null,
13 "tool_calls": null
14 }
15 }
16 ],
17 "created": 1741224586,
18 "model": "",
19 "object": "chat.completion",
20 "service_tier": null,
21 "system_fingerprint": null,
22 "usage": {
23 "completion_tokens": 145,
24 "prompt_tokens": 38,
25 "total_tokens": 183,
26 "completion_tokens_details": null,
27 "prompt_tokens_details": null
28 }
29}
