Google: Gemma 4 31B (free)

Freegoogle/gemma-4-31b-it:free

Input / M
$0
Output / M
$0
Context
262.1K
Max output
32.8K
Listed since
Apr 2, 2026
Can do
Tools · Vision · Reasoning

Reliability

Measured · signed cheapoS reports
8.3%

answered
likely 2.3–25.8%
2 of 24 outcomes

Avg time
6.9 s
24 timed
Failures
22
22 rate-limited
Reports
24
0 cancelled or unknown
Members
1
early sample

Sep 20, 2026 to Oct 5, 2026 · OpenRouter via OmniRoute. The range is a 95% interval and narrows as more members report. “Answered” means the provider responded, not that the task succeeded.

Failure breakdown
  • rate limit quota: 22

Where to get it

All free tiers →
ListingPrice / MLimitsLink
OpenRouter · free routegoogle/gemma-4-31b-it:free$0Free model routes (IDs ending in :free) are limited to 20 requests per minute. Free routes allow 50 requests per day, or 1,000 per day once the account has bought at least $10 in credits.Open ↗
OpenRouter · paid routegoogle/gemma-4-31b-it$0.20 / $0.40Pay as you go; no free-route caps.Open ↗

Free routes are served by different providers with their own data policies. Check a route's page on OpenRouter before sending private code. OpenRouter terms, checked Sep 29, 2026

About

OpenRouter description

Gemma 4 31B Instruct is Google DeepMind's 30.7B dense multimodal model supporting text and image input with text output. Features a 256K token context window, configurable thinking/reasoning mode, native function...

Knowledge cutoff: Not reported.

Catalog data from the OpenRouter model catalog, retrieved Nov 16, 2026, 5:53 PM UTC. A listing isn't a live availability check. How we measure