UTC

Gemma 4 31B IT

Google · open weights · served by 58 providers · tool calling, reasoning, image input

google/gemma-4-31b-it

Weights
Open
Released
2026-04-02
Knowledge cutoff
2025-01
Context window
262K
Deprecation dates

When each platform deprecates Gemma 4 31B IT, as LiteLLM records it. Plan the move before the date: after it, the platform may stop serving the model or stop updating it.

  • Together AIdeprecated 2026-09-15together_ai/google/gemma-4-31B-itsuccessor: together_ai/zai-org/GLM-5.3-Flash
Every upcoming deprecation →
Hugging Face repository

google/gemma-4-31B-it at commit 842da3794e, last changed 2026-07-20

Licence
apache-2.0
Access
Open download
Library
transformers
Formats
.safetensors
  • No model signature file is published with the weights.
  • Built on google/gemma-4-31B.

What the Hub's scanners report

2 weight files, 2 with a scan result. No scanner reports an issue with them.

The model at this commit, its 2 weight files with their SHA-256 and its base models.

Facts from the Hugging Face Hub and its file scanners (Protect AI, Cisco, Hugging Face picklescan, VirusTotal, JFrog), as of 2026-11-18. ShieldMap does not copy model cards.

Who serves it
  • Abacusgoogle/gemma-4-31b-it
  • Adaptiongemma-4-31b-it
  • ai&google/gemma-4-31b-it
  • AIHubMixgemma-4-31b-it-free
  • AI/ML APIgoogle/gemma-4-31b-it
  • Amazon Bedrockgoogle.gemma-4-31b
  • AMDGemma-4-31B-IT
  • Berget.AIgoogle/gemma-4-31B-it
  • Bothubgemma-4-31b-it:free
  • Chutesgoogle/gemma-4-31B-turbo-TEE
  • Cortecsgemma-4-31b-it
  • Crusoegoogle/gemma-4-31b-it
  • Deep Infragoogle/gemma-4-31B-it
  • DigitalOceangemma-4-31B-it
  • Doublewordgoogle/gemma-4-31B-it
  • FastRoutergoogle/gemma-4-31b-it
  • Friendligoogle/gemma-4-31B-it
  • Hugging Facegoogle/gemma-4-31B-it
  • InferXgemma-4-31B-it-fp8
  • Infomaniakgoogle/gemma-4-31B-it
  • Kelngoogle/gemma-4-31b-it
  • Kilo Gatewaygoogle/gemma-4-31b-it
  • DevPass (LLM Gateway)gemma-4-31b-it
  • LLM Gatewayconsensusprotocol/gemma-4-31b-it
34 more providers

LLM Gateway (deepinfra/gemma-4-31b-it), LLM Gateway (novita/gemma-4-31b-it), LLM Gateway (runware/gemma-4-31b-it), LLMTR (gemma-4), Merge Gateway (google/gemma-4-31b-it), Modular (google/gemma-4-31b-it), Moneo (gemma-4-31b:web), NanoGPT (TEE/gemma-4-31b-it), NanoGPT (TEE/gemma4-31b), NanoGPT (TEE/gemma4-31b:thinking), NanoGPT (google/gemma-4-31b-it), NanoGPT (google/gemma-4-31b-it:thinking), Neuralwatt (gemma-4-31b), Novita AI (google/gemma-4-31b-it), Nvidia (google/gemma-4-31b-it), Ollama Cloud (gemma4:31b), OpenRouter (google/gemma-4-31b-it), OpenRouter (google/gemma-4-31b-it:free), Opper (gemma-4-31b-it), OrcaRouter (google/gemma-4-31b-it), Phala (google/gemma-4-31b-it), Poe (google/gemma-4-31b), Powerspring (gemma-4-31b), QVAC (gemma4-31b), Regolo AI (gemma4-31b), Requesty (gemma-4-31b-it), SambaNova (gemma-4-31B-it), SiliconFlow (google/gemma-4-31B-it), stepping stone AG (google/gemma-4-31B-it), Tinfoil (gemma4-31b), UnoRouter (gemma-4-31b-it:free), Venice AI (google-gemma-4-31b-it), Vercel AI Gateway (google/gemma-4-31b-it), CoreWeave (google/gemma-4-31B-it)

Running it yourself

Open weights are loaded and served by code that has its own advisories. These inference servers and model loaders are on ShieldMap; load weights only from the publisher's own repository, and prefer safetensors over formats that can run code.

How supply-chain attacks on models work →

Sources: models.dev (MIT) and the LiteLLM model catalogue (MIT). Deprecation dates are as each platform published them and as LiteLLM records them; check the platform before you rely on one. Data as of 2026-11-22.