Qwen3 Coder Next
alibaba/qwen3-coder-next- Alibaba
- Family · Qwen3 Coder Next
- Released 3 Feb 2026
- Open weights
- Text → Text
- Tools · JSON
Spec & list price
Context256K
Max output236K
Input / 1M$0.12
Output / 1M$0.806.7× input
Scores · standing
Intelligence9.2#266/489
Coding36.2#101/196
Usage
Tokens / day3.1B▲ 13% 7d
Usage rank#141
Requests / day60.9K
Tool calls / req0.54
Per request · 30 days
Avg input31.1K2.1× median
Avg output3440.5× median
Cache hit88%1.3× median
Tool-call failures8.3%4.4× median
Consumption
Peak 7.3B 10 OctLow 1.7B 26 SepRank today #141
News & trends
Trending · 14 days
- Qwen3-Coder-Next en PC de Escritorio: ¿El Rey del Código Local o Ilusión 2D? (80B MoE en CPU)94 views
News · latest from Alibaba
- Memdoor: A team of 5 Qwen3.8-27B models beats Fable 5 on LiveCodeBench Hardgithub.com · 31 Oct
- Qwen3.8-Flash Now Available in the EU on Alibaba Cloudtwitter.com · 31 Oct
- Alibaba's Qwen3.8-Flash-Next model tried to upload a file to Beijing in a testnotebookcheck.net · 27 Oct
- Running Qwen3.8-Flash-Next NVFP4 at 184 tok/s on a single RTX 6000 Promeshapplied.com · 26 Oct
Benchmarks
bar = share of the top score · Benchmark data: Artificial AnalysisSpeed · median across providers, latest day
Output speed#185/33460 tok/sabout median
First token#1120.74 s1.4× quicker than median
#16 cheapest of 41 with a similar score
Value
Intelligence Index
Blended price per 1M tokens (log)
Technical details
- Tokenizer
- Qwen
- License
- Apache 2.0
- Parameters
- 79.7B
- Moderation
- none
- Downloads · likes
- 434.1K · 1.7K
Request parameters 16
- tools
- tool_choice
- response_format
- structured_outputs
- temperature
- top_p
- top_k
- max_tokens
- stop
- seed
+6 more
- frequency_penalty
- logit_bias
- logprobs
- presence_penalty
- repetition_penalty
- top_logprobs
When to use
Best for
- Coding assistants with tool calling
- Agentic terminal and repository tasks
- Self-hosted code generation
- Structured output for developer tooling
Think twice for
- Image understanding (text-only input)
- Hardest autonomous coding runs against top models
- General creative writing
Research papers
- VulContextBench: A Benchmark for Security Context Retrieval in Coding Agents26 Sep
- From Execution to Education: A Bloom-Aligned Framework for Measuring Educational Control in LLMs9 Jul
- An Empirical Study of Security Calibration in Large Language Models for Code30 Jun
- Generated, Parallel, Scalable? A Study of Agentic AI-Generated Julia Code on Supercomputers15 Jun
- Security in a Workflow: Exploring Role-Based Agentic Architectures for Vulnerability Handling12 Jun
- PDE-Agents: An LLM-Orchestrated Multi-Agent Framework for Automated Finite Element Simulations with Knowledge Graph-Augmented Reasoning5 Jun
- QBugLM: An Agentic Benchmarking Framework for LLM-based Quantum Software Debugging5 Jun
- Qwen3-Coder-Next Technical Report28 Feb
Social
Most watched and discussed · 90 days
- Qwen3-Coder-Next en PC de Escritorio: ¿El Rey del Código Local o Ilusión 2D? (80B MoE en CPU)Christian Leyva · 94 views · 19 Oct
Alternatives
other vendors · similar capability- 12.9
- 8.5
- 11.6
- 13.1
Using Qwen3 Coder Next at volume?
Sign in to ask for a volume discount. You will continue in the OSMEXO dashboard.