Alibaba QwenProvisional

Qwen3-Coder-Next

Qwen / Qwen3-Coder-Next · Qwen/Qwen3-Coder-Next

Efficient open-weight coding agent model for local and cost-sensitive development workflows.

Objective score

17.2

Confidence

79%

Price evidence

Pricing evidence pending

Support this model Compare trend Official source

Other primary tiers in this family

Qwen3-235B-A22B Instruct 2507 Qwen3-235B-A22B Thinking 2507 Qwen3-Coder 480B-A35BQwen3-Coder-Next

Scenario evidence

Six task-specific rankings

Feature development#65Obj. 17.20 · 0%Debug and repair#61Obj. 27.10 · 0%Refactoring#61Obj. 27.10 · 0%Code review#61Obj. 28.10 · 0%Large-repo agent work#65Obj. 14.60 · 0%Frontend / UI#61Obj. 27.10 · 0%

Evidence state

Signal coverage

Benchmarks: 2
Rank points: 8
Usage: 2
Agent: 0
Sources: 3

Objective evidence

Benchmarks and public signals

Swebench: —
Livecodebench: —
Aider: —
Artificial Analysis: 36.2
Openrouter Usage: 0.1
Kilo Usage: —

Persistent history

Recent ranking points

Aug 1#64 · 18

Aug 1#64 · 18

Jul 31#64 · 18

Jul 31#64 · 18

Jul 31#64 · 18

Jul 31#64 · 18

Usage signals

OpenRouter0.1%

Overall0.1%

Agent and tool signals

Evidence is being collected. Missing data is not treated as zero.

Source trail

Official Qwen3-Coder repo describes Coder-Next, 256K native context, and coding-agent positioning.Artificial Analysis Coding Agents · ok Kilo leaderboard · ok Qwen3-Coder repository · ok