Qwen3.5-27B
alibaba/qwen3.5-27b- Alibaba
- Family · Qwen3
- Released 24 Feb 2026
- Open weights
- Text · Image · Video → Text
- Tools · JSON · Vision
Not on OSMEXO yet
Spec & list price
Context256K
Max output80K
Input / 1M$0.26
Output / 1M$2.6010× input
Scores · standing
Intelligence22.9#111/490
Usage
Tokens / day284.0M▼ 32% 7d
Usage rank#247
Requests / day107.2K
Tool calls / req0.12
Per request · 30 days
Avg input3.9K0.3× median
Avg output4140.6× median
Cache hit2.4%0.0× median
Reasoning share62%about median
Tool-call failures3.3%1.7× median
Consumption
Peak 4.8B 19 OctLow 284.0M 1 NovRank today #247
News & trends
2 mentions this week ▲ 100%this model · videos, posts, articles · 30 days
Trending · 14 days · Qwen3 family
- Qwen3.8 27b (Q6) casually breaking reCAPTCHA838 upvotes
- Memdoor: A team of 5 Qwen3.8-27B models beats Fable 5 on LiveCodeBench Hard416 points
- Neuron Explorer for Qwen3.5-4B: See every neuron at work, make targeted edits26 points
- New Card for Qwen3.8 27b and MiniMax H3656 upvotes
- Qwen 3.8 27B vs Strata Coder - Local LLM battle75.2K views
- DavidAU Qwen3.8 27B Turbo Brilliance 50X-Reasoning... tested - Local LLM setup87.7K views
News · latest from Qwen3
- Memdoor: A team of 5 Qwen3.8-27B models beats Fable 5 on LiveCodeBench Hardgithub.com · 31 Oct
- Neuron Explorer for Qwen3.5-4B: See every neuron at work, make targeted editsgithub.com · 26 Oct
- How to train your own Jev for $17Together AI · 23 Sep
Benchmarks
bar = share of the top score · Benchmark data: Artificial AnalysisSpeed · median across 6 providers, latest day
Output speed#283/33435 tok/s1.8× slower than median
Fastest provider65 tok/sbest of 6 providers
First token#2221.6 s1.5× longer than median
#16 cheapest of 32 with a similar score
Value
Intelligence Index
Blended price per 1M tokens (log)
Family progress
Intelligence Index
- Family releases33.7
Technical details
- Tokenizer
- Qwen3
- License
- Apache 2.0
- Parameters
- 27.8B
- Moderation
- none
- Downloads · likes
- 1.8M · 1.1K
Request parameters 19
- tools
- tool_choice
- response_format
- structured_outputs
- reasoning
- include_reasoning
- temperature
- top_p
- top_k
- min_p
+9 more
- max_tokens
- stop
- seed
- frequency_penalty
- logit_bias
- logprobs
- presence_penalty
- repetition_penalty
- top_logprobs
When to use
Best for
- Tool-calling agents on a modest compute budget
- Strict instruction following in structured pipelines
- Long-context reasoning over documents
- Image and video understanding
- Self-hosted assistants (open weights)
Think twice for
- Image or video generation
- Hardest terminal-based coding agents
- Audio tasks
- Very long single outputs
Research papers
- ThinkJev: One Thought, Every Budget for Calibrated Typed Decisions27 Oct
- Learning to Formalize: Bridging Open-Ended Workloads and Fast Decision Models27 Oct
- Symbolic Constraint Agents: Mining Formal Tool-Call Constraints to Prevent Agent Errors18 Oct
- When Protection Overrides Consequences: Understanding Protection Mode in Large Language Model Decision-Making14 Oct
- BANGLAPLAN: A Benchmark for Constraint-Based Planning and Decision-Making in Bangla14 Oct
- Artifact Distillation: Training-Free Transfer from Frontier to Small Language Models12 Oct
- StoreBench: A Live-Commerce Environment for Evaluating and Training Autonomous Operator Agents7 Oct
- Learning from Research: Toward Lifelong Agent Harness Evolution30 Sep
Social
Most watched and discussed · 90 days
- Qwen3.5 27B vs 9B on a 12GB GPU or 18GB Mac: Skip the 2-BitCode Craft Studio · 14 views · 23 Oct
Same family
Qwen3- 7.2
- 6.0
- 6.9
- 6.1
- 5.2
- 13.1
Alternatives
other vendors · similar capability- 26.5
- 26.3
- U26.6
- 21.8
Want Qwen3.5-27B on OSMEXO?
Sign in to request it. Requests are reviewed; a request does not guarantee availability.