local

unsloth/gemma-4-12b-it-GGUF

Released Context 0Kunsloth-gemma-4-12b-it-gguf

PipelineScore

77.5MAINLINE

Ranked #15 of 23 models · 35th percentileRAG is the headline (100.0); throughput is the soft spot (42.5). Best-fit profile: Agentic.

Category breakdown

Score per category, normalized 0–100 against the v1 anchor.

Code

81.3

Reason

70.0

Tool Use

87.5

RAG

100.0

Speed

42.5

Strengths

RAG100.0

Tool Use87.5

Code81.3

Same model, different rigs

Every submission of unsloth/gemma-4-12b-it-GGUF on the 0–100 scale. The spread is the point: where it runs changes what you get.

0255075100

Best 83.2 on rx9070xt-16gb-i7-6850k-32gb-ddr4-2400 · lowest 73.0 on rx9070xt-16gb-i7-6850k-32gb-ddr4-2400 · spread 10.1 pts across 4 runs. Hover a dot for its rig.

Sample tasks

A taste of what the test pack measures. Full prompts are private and rotated daily.

CodeDifficulty 1code-fib-1

Fibonacci function

Write a Python `fib(n)` returning the nth Fibonacci number, O(n).

ReasonDifficulty 1reason-math-1

Train meeting time

Two trains, opposite directions, given speeds and start times — when do they meet?

RAGDifficulty 2rag-extract-1

Extract metrics to JSON

From the context, extract net sales, operating margin, and free cash flow as a JSON object. Numbers only.

Tool UseDifficulty 2tool-schema-1

OpenAPI param selection

Given an OpenAPI schema with limit/offset/sort, fill JSON for 'next 50, recent first.'

RAGDifficulty 2rag-grounding-1

Refuses to fabricate

Context lacks the answer — does the model fabricate or correctly say it can't?

Compare with

unsloth/gemma-4-12b-it-GGUF vs openai/gpt-oss-20b unsloth/gemma-4-12b-it-GGUF vs gpt-oss:latest unsloth/gemma-4-12b-it-GGUF vs unsloth/gpt-oss-20b-GGUF unsloth/gemma-4-12b-it-GGUF vs unsloth/gemma-4-E4B-it-GGUF unsloth/gemma-4-12b-it-GGUF vs Gemma