modelbenchmark.io

Qwen3.6 27B

Alibaba · qwen3-6

Compare

Qwen vision-language model for visual reasoning, documents, and agent tasks

Specification

most-agreed values

Context
262K
Max output
66K
Released
2026-04-22
Knowledge cutoff
Retires
Open weights
yes
Input
text, image
Output
text

Price

US dollars per million tokens · most-agreed

Input
free
Output
free

Quality

7 benchmarks · 1 source

Composite
83rd
percentile of 334 scored models
Rank
59 / 334
±0.06 sd
Evidence
7 × 1
one source, or fewer than 3 benchmarks
Effort range
one configuration only
coding86th1/3 bench
reasoning84th2/4 bench
math76th3/7 bench
knowledge76th1/1 bench

Every score

one row per source and configuration — nothing averaged away

BenchmarkScoreConfigurationSourceRun
OTIS Mock AIME 2024-202591.1 ±4.3maxEpoch AI2026-08-07
GPQA diamond87.4 ±2.4maxEpoch AI2026-08-07
SWE-Bench verified76.7 ±1.9maxEpoch AI2026-05-28
SimpleQA Verified52.0 ±1.6maxEpoch AI2026-08-27
FrontierMath-v136.5 ±2.5max · 2 runsEpoch AI2026-05-27
Chess Puzzles20.0 ±4.0maxEpoch AI2026-08-07
FrontierMath-Tier-4-2025-07-012.1 ±2.9max · 2 runsEpoch AI2026-05-28

The leading figure for a benchmark is the median across its configurations, so one heroic high-effort run cannot set the number.

Available from 2 hosts

HostIn $/MOut $/MCache rdCache wrContextOutputRetires
nanfreefree262K66K
pendra · 27bfreefree262K66K

More from Alibaba

most-hosted first