No longer supported

This model is deprecated (released May 2024) and is no longer actively maintained or recommended by Google. Consider using their current flagship model instead.

deprecatedlegacyllm Proprietary

PaLM 2 Otter

by Google· Released May 2024

PaLM 2 Otter is a lightweight, fast variant of Google's PaLM 2 model family, optimized for high-throughput and low-latency tasks. It was designed for simple text generation and classification, but has been deprecated in favor of newer Gemini models.

Official Site

Input cost

—

Output cost

—

Context window

8K tokens

Max output

2048 tokens

Modalities

text

License

proprietary

Capabilities

Text GenerationClassificationSentiment AnalysisEntity Extraction

Best For

Simple text generation and classification tasks requiring low latency and high throughput.

Strengths

Fast inference speed
Low cost per token
Efficient for simple tasks
Good for high-volume applications

Limitations

Limited context window (8K tokens)
No multimodal capabilities
Outperformed by Gemini models
Deprecated and no longer recommended

Use Cases

Chatbots for simple Q&A

Text classification

Sentiment analysis

Entity extraction

Content moderation

Simple summarization

Language translation

Back to all models