google

Google: Lyria 3 Clip Preview

Google: Lyria 3 Clip Preview is a specialized audio generation model from google, available at $0 per million tokens during the preview period. It accepts text and image inputs and produces text and audio outputs, with a 1,048,576-token context window — the same architecture as Lyria 3 Pro Preview. As an audio generation model, it operates outside the domain of our standard 12-benchmark LLM testing suite, which focuses on text-based capabilities. We do not have benchmark scores or rankings for this model. The 'Clip' designation suggests it is optimized for shorter-form or clip-length audio outputs, though the payload contains no description to confirm this.

Performance

Google: Lyria 3 Clip Preview has not been included in our standard 12-benchmark testing suite. Our benchmarks evaluate text-based LLM capabilities — classification, agentic planning, faithfulness, constrained rewriting, and similar dimensions — which do not apply to audio generation models. We cannot report scores or rankings for this model. Teams evaluating Lyria 3 Clip Preview should assess it directly on audio generation quality for their target use cases.

Pricing

Lyria 3 Clip Preview is currently priced at $0 per million tokens for both input and output — free to access during the preview period. Like all preview-stage models, pricing and availability terms may change upon general availability. For teams wanting to explore audio generation from google at no cost, the preview window provides a zero-risk evaluation opportunity. No usage-volume cost estimates are relevant at the current $0 rate.

google

Google: Lyria 3 Clip Preview

Overall
0.00/5N/A

External Benchmarks

SWE-bench Verified
N/A
MATH Level 5
N/A
AIME 2025
N/A

Pricing

Input

$0.00/MTok

Output

$0.00/MTok

Context Window1049K

modelpicker.net

Real-World Costs

iChat response$0.00
iBlog post$0.00
iDocument batch$0.00
iPipeline run$0.00

Pricing vs Performance

Output cost per million tokens (log scale) vs average score across our 12 internal benchmarks

This modelOther models

Try It

from openai import OpenAI

client = OpenAI(
    base_url="https://openrouter.ai/api/v1",
    api_key="YOUR_OPENROUTER_KEY",
)

response = client.chat.completions.create(
    model="google/lyria-3-clip-preview",
    messages=[
        {"role": "user", "content": "Hello, Google: Lyria 3 Clip Preview!"}
    ],
)

print(response.choices[0].message.content)

Recommendation

Google: Lyria 3 Clip Preview is relevant for developers and teams building audio generation applications who want to access google's audio model at no cost during the preview period. The 'Clip' variant likely targets shorter audio segments compared to the Pro variant, making it a candidate for short-form audio generation use cases such as sound clips, short music segments, or brief audio samples. Since this falls outside our LLM benchmarking scope, use case fit should be validated with direct testing. If you need a text-based model from google, Gemma 4 31B (avg 4.416 in our testing) or Gemini 3.1 Pro Preview (avg 4.333) are stronger documented options.

How We Test

We test every model against our 12-benchmark suite covering tool calling, agentic planning, creative problem solving, safety calibration, and more. Each test is scored 1–5 by an LLM judge. Read our full methodology.