Multimodal
inferx
merge-gateway
google
nano-gpt
nvidia
openrouter
Gemma Family
Open Source
Released: Apr 2026
Updated: Aug 2026 Gemma 4 31B It
Gemma 4 31B It is a Multimodal model.
Largest Gemma 4 instruction model for open, self-hosted chat and reasoning
Vision
Audio
Tool Use
Reasoning
Citations
Key Specs
Context
262k
Max Output
262k
Best Price (Input / Output)
Chat with Model
$0.10 / $0.34 per 1M tokens
Pricing Comparison
| Provider | Input (1M) | Output (1M) | Image (1k) |
|---|---|---|---|
| inferx (FP8) | Free | Free | - |
| nvidia | Free | Free | - |
| openrouter (Free Tier) | Free | Free | - |
| openrouter | $0.10 | $0.34 | - |
| merge-gateway | $0.14 | $0.40 | - |
| nano-gpt (TEE) | $0.15 | $0.46 | - |
Prices are per 1 million tokens unless otherwise noted. Image pricing is per 1000 images if applicable.
Price History
Prompt Price
Completion Price
Price trend per 1 million tokens over time.
Supported Parameters
frequency_penalty
reasoning
min_p
structured_outputs
response_format
include_reasoning
stop
top_k
max_tokens
top_logprobs
temperature
presence_penalty
tools
seed
tool_choice
logit_bias
logprobs
repetition_penalty
top_p
top_a