Poe API
Gemma-4-31B
By
Google's most capable open model, delivering frontier-class reasoning and multimodal understanding in a 31 billion parameter dense architecture. Gemma 4 supports a 256K token context window, understands over 140 languages, and accepts both text and image inputs. It features built-in chain-of-thought reasoning for complex multi-step tasks, strong coding and math performance, and function calling support. Released under the Apache 2.0 license, Gemma 4 31B offers competitive performance with leading proprietary models while being fully open-weight. This bot may experience degraded performance due to high traffic to this model.
Powered by Google: gemma-4-31b-it. Learn more
- OFFICIAL
Build with Gemma-4-31B using the Poe API
Start by creating an API key, for use with any bot on Poe:
Use it in these bots
See allCode samples
Responses API
import openai
client = openai.OpenAI(
api_key = "YOUR_POE_API_KEY", # or os.getenv("POE_API_KEY")
base_url = "https://api.poe.com/v1",
)
response = client.responses.create(
model = "gemma-4-31b",
input = "Translate 'The world is full of beauty.' into French"
)
print(response.output_text)See the full documentation for comprehensive guidance on getting started.
More from Google
Gemma-4-31B
Veo-3.1-Lite
Lyria-3
Nano-Banana-2
Gemini-3.1-Pro
Gemini-3.1-Flash-Lite
Nano-Banana-Pro
Veo-3.1-Fast
Veo-3.1
Veo-3-Fast







