← Back to leaderboard
local
Jackrong/Qwen3.5-9B-DeepSeek-V4-Flash-GGUF
Released Context 0Kjackrong-qwen3.5-9b-deepseek-v4-flash-gguf
PipelineScore
79.3MAINLINERanked #19 of 24 models · 21st percentileCode is the headline (100.0); throughput is the soft spot (35.8). Best-fit profile: Coding.
Category breakdown
Score per category, normalized 0–100 against the v1 anchor.
Code
100.0
Reason
60.0
Tool Use
87.5
RAG
100.0
Speed
35.8
Strengths
Code100.0
RAG100.0
Tool Use87.5
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
Jackrong/Qwen3.5-9B-DeepSeek-V4-Flash-GGUF vs openai/gpt-oss-20bJackrong/Qwen3.5-9B-DeepSeek-V4-Flash-GGUF vs gpt-oss:latestJackrong/Qwen3.5-9B-DeepSeek-V4-Flash-GGUF vs unsloth/gpt-oss-20b-GGUFJackrong/Qwen3.5-9B-DeepSeek-V4-Flash-GGUF vs unsloth/gemma-4-E4B-it-GGUFJackrong/Qwen3.5-9B-DeepSeek-V4-Flash-GGUF vs Gemma