Qwen3.5 27B (Reasoning)
AlibabaQwenOpen WeightApache 2.0 · Commercial OK
Description
Qwen3.5-27B is a multimodal dense foundation model with 27 billion parameters. It combines strong reasoning, coding, multilingual, long-context, and visual understanding performance in a production-friendly open-weight package with a native 262K context window.
Release Date
2026-02-24
Parameters
27.0B
Context Length
262K
Modalities
image, text, video
Capability Radar
70
general
60
coding
80
reasoning
77
scienceest.
60
agents
80
multimodal
Science is estimated from LLM Stats science scores or reasoning when no dedicated science benchmarks are available.
Rankings
| Domain | #Rank | Score | Source |
|---|---|---|---|
| Agentic Capability | 66 | 49.0 | LS |
| Multimodal Ranking | 72 | 53.0 | LS |
Benchmark Scores (LLM Stats)
(LLM Stats (zeroeval))Agents
t2-bench
79.0%SR
VITA-Bench
41.9%SR
DeepPlanning
22.6%SR
Chat
IFEvalGoogle Research (2023)
95.0%SR
Multi-Challenge
60.8%SR
Code
FullStackBench en
60.1%SR
FullStackBench zh
57.4%SR
General
C-Eval
90.5%SR
MAXIFE
88.0%SR
Include
81.6%SR
AndroidWorld_SR
64.2%SR
NOVA-63
58.1%SR
Healthcare
MedXpertQA
62.4%SR
PMC-VQA
62.4%SR
Instruction Following
IFBench
76.5%SR
Language
MMLU-Redux
93.2%SR
MMLU-Pro
86.1%SR
MMMLU
85.9%SR
MMLU-ProX
82.2%SR
WMT24++
77.6%SR
Long Context
LongBench v2
60.6%SR
Math
HMMT 2025
92.0%SR
HMMT25
89.8%SR
MathVista-Mini
87.8%SR
DynaMath
87.7%SR
MathVision
86.0%SR
CodeForces
0.81 / 3000SR
PolyMATH
71.2%SR
Multimodal
VideoMME w/o sub.
82.8%SR
VideoMMMU
82.3%SR
MMMU
82.3%SR
TIR-Bench
59.8%SR
OSWorld-Verified
56.2%SR
Reasoning
Global PIQA
87.5%SR
GPQANYU + Cohere + Anthropic (2023)
85.5%SR
LiveCodeBench v6
80.7%SR
CharXiv-R
79.5%SR
SWE-Bench Verified
72.4%SR
AA-LCR
66.1%SR
SuperGPQA
65.6%SR
BrowseComp-zh
62.1%SR
BrowseCompOpenAI (2025)
61.0%SR
Humanity's Last Exam
48.5%SR
Seal-0
47.2%SR
Terminal-Bench 2.0Stanford × Laude Institute (2026)
41.6%SR
OJBench
40.1%SR
Search
WideSearch
61.1%SR
Tool Calling
BFCL-V4
68.5%SR
Video
MLVU
85.9%SR
Vision
CountBench
0.98 / 100SR
VLMsAreBlind
96.9%SR
V*
93.7%SR
AI2D
92.9%SR
MMBench-V1.1
92.6%SR
RefCOCO-avg
0.91 / 100SR
OCRBench
89.4%SR
OmniDocBench 1.5
88.9%SR
VideoMME w sub.
87.0%SR
EmbSpatialBench
0.84 / 100SR
RealWorldQA
83.7%SR
LingoQA
82.0%SR
CC-OCR
81.0%SR
MMStar
81.0%SR
SlakeVQA
80.0%SR
MMMU-Pro
75.0%SR
MVBench
74.6%SR
LVBench
73.6%SR
MMVU
73.3%SR
ScreenSpot Pro
70.3%SR
Hallusion Bench
70.0%SR
RefSpatialBench
0.68 / 100SR
ERQA
60.5%SR
MMLongBench-Doc
0.60 / 100SR
SimpleVQA
0.56 / 100SR
BabyVision
44.6%SR
ODinW
41.1%SR
ZEROBench-Sub
0.36 / 100SR
SUNRGBD
0.35 / 100SR
Nuscene
15.2%SR
Hypersim
0.13 / 100SR
ZEROBench
0.10 / 100SR
AA Evaluation Indices
(Artificial Analysis)No AA evaluation data available
LLM Stats Category Scores
(LLM Stats (zeroeval))Instruction Following90
Biology90
Chat80
Image To Text80
Language80
Legal80
Math80
Physics80
Structured Output80
Embodied80
Finance80
Grounding80
Chemistry80
Text-to-image80
Video80
Long Context70
Multimodal70
Reasoning70
Spatial Reasoning70
Frontend Development70
General70
Healthcare70
Economics70
Vision70
Search60
Agents60
Code60
Communication60
Tool Calling60
Spatial20
3d20
Pricing
Input Price$0.3 / 1M tokens
Output Price$2.4 / 1M tokens
Blended Price (3:1)$0.795 / 1M tokens
Speed
Tokens/sec29.0
Time to First Token5.55s
Time to Answer74.58s
Provider Price Ranking
Provider Price Ranking
4 providers
Cheapest: EmpirioLabs AIMost Expensive: Novita
ProviderInputOutput
1EmpirioLabs AICheapest
$0.086
$0.688
2DeepInfra
$0.26
$2.6
3AlibabaPRIMARY
$0.3
$2.4
4Novita
$0.3
$2.4
Compare pricing across different API providers for this model.