gemma4

Gemma 4 models are designed to deliver frontier-level performance at each size. They are well-suited for reasoning, agentic workflows, coding, and multimodal understanding.

vision tools thinking audio cloud e2b e4b 12b 26b 31b

19M Pulls 49 Tags Updated 2 weeks ago

medgemma1.5

MedGemma 1.5 4B is an updated version of the MedGemma 4B model.

vision 4b

87.6K Pulls 5 Tags Updated 3 months ago

4skl/gemma4-12b-mtp

High-tier 12B Gemma 4 model with Unsloth Dynamic QAT & MTP. Packs deep multimodal reasoning, an aggressive 4-token speculative draft window, and a massive 128K context capacity. Tailored for heavy codebase ingestion and reasoning tasks.

tools thinking

183 Pulls 1 Tag Updated 3 days ago

4skl/gemma4-e4b-mtp

Multimodal 4.5B Gemma 4 with Unsloth Dynamic QAT & MTP (5.3GB). Features near-lossless vision, audio, text reasoning, an active 4-token speculative window, and an expanded 64K context size. Best for edge execution on systems with 8GB to 16GB RAM.

vision tools thinking

156 Pulls 1 Tag Updated yesterday

4skl/gemma4-e2b-mtp

Ultra-fast multimodal 2.3B Gemma 4 for on-device edge AI (3.7GB). Adds native vision/audio parsing to Unsloth Dynamic QAT with a 2-token MTP pipeline and a mobile-safe 32K context window. Perfect for smartphones and laptops sharing 8GB of total RAM.

vision tools thinking

104 Pulls 1 Tag Updated yesterday

byczech/gemma4-it-mmproj-16G-Q3_K_S

Gemma 4 31B multimodal instruct model with vision support, quantized to Q3_K_S and optimized for 16 GB VRAM. Supports a practical context window of approximately 36K–40K tokens, depending on the Ollama version, GPU, backend and runtime configuration.

vision tools thinking 31b

25 Pulls 1 Tag Updated 2 days ago

imranzunzani/gemma4-26bitqat-owui-fix

Fixes the issue with Open WebUI not passing tools (native) and system prompt information properly. It requires bigger value for num_ctx default: 32768. Tested with Open WebUI v0.9.6 and v0.10.2. Just built on top of official qat of gemma4-26b

vision tools thinking

6 Pulls 1 Tag Updated 5 days ago

johnblick187/gemma4-rp

vision tools thinking

4 Pulls 1 Tag Updated 5 days ago

vinkas/VinkasGemma4

vision tools thinking

2 Pulls 1 Tag Updated yesterday

ecloudsenergyllp/gemma4

vision tools thinking

1 Pull 1 Tag Updated 19 hours ago

satgeze/gemma4-12b-uncensored-1.5m

Gemma 4 12B uncensored, 1.5M max context: certified 10/10 to 1.31M on a 32GB card. Vision included.

vision tools thinking

1,712 Pulls 1 Tag Updated 2 weeks ago

baytout3/gemma4-12b-qat-uncensored-hauhaucs-balanced

tools thinking

2,359 Pulls 1 Tag Updated 3 weeks ago

satgeze/gemma4-26b-uncensored-1m

Gemma 4 26B-A4B uncensored MoE: 1M context, vision, fast. ~91% recall, honestly documented.

vision tools thinking

1,015 Pulls 1 Tag Updated 2 weeks ago

satgeze/gemma4-12b-uncensored-1m

Gemma 4 12B uncensored: first Gemma certified 10/10 at 1M, fits a 32GB card. Vision included.

vision tools thinking

548 Pulls 1 Tag Updated 2 weeks ago

pierreprudh/gemma4-12b-mtp

Gemma 4 12B (QAT Q4_0) with Multi-Token Prediction speculative decoding — a fast, tool-capable local model for 16 GB - Apple Silicon Macs (~21 tok/s, ~1.4× via MTP).

tools thinking

563 Pulls 1 Tag Updated 3 weeks ago

MobiusDevelopment/gemma4-E2B-it-qat-Q4-unsloth-heretic

Gemma 4 models are designed to deliver frontier-level performance at each size. They are well-suited for reasoning, agentic workflows, coding, and multimodal understanding.

vision tools thinking

543 Pulls 1 Tag Updated 3 weeks ago

artokun/gemma4-comfyui-mcp

Gemma 4, fine-tuned to drive ComfyUI. A size ladder of Google's Gemma 4 models QLoRA-trained on 1,055 server-verified tool-use trajectories generated against a live ComfyUI instance — covering the complete comfyui-mcp tool surface

tools thinking e2b e4b 12b

264 Pulls 3 Tags Updated 1 week ago

edtorre/gemma4

Gemma 4 12B agent-tuned for Hermes Agent and other local agent harnesses. One built for 20GB GPUs the other for 8GB GPUs.

vision tools thinking

230 Pulls 2 Tags Updated 1 week ago

sonct988/gemma4-26b-a4b-it-q4km-256k

Gemma 4 26B A4B Instruct GGUF Q4_K_M build for Ollama, configured with a 256K context window. This is a text-focused local model built from google/gemma-4-26B-A4B-it and intended for chat, summarization, tool-style workflows, and long-context testing.

tools thinking

116.6K Pulls 1 Tag Updated 1 month ago

edtorre/gemma4-26b-a4b-it-qat-agent

Gemma 4 26B tuned for agentic use. 64K context window, flash attention + Q8 KV cache quantization for reduced VRAM. Temperature 0.7, output capped at 8192 tokens.

vision tools thinking

131 Pulls 1 Tag Updated 2 weeks ago