Qwen3 Next 80B A3B Thinking
Qwen released 2025-09-11
DeepSeek V4 Flash 0731 is both better and cheaper.
It scores +34.9 higher and costs 75% less ($0.105/M against $0.412/M) on this workload — and it does everything this model does.
Gemma 4 26B A4B is cheaper still (67% less) but drops 33K → 16K max output.
Every price dimension
| Input | $0.150/M |
|---|---|
| Output | $1.20/M |
Independent scores
| Intelligence | 16.9 |
|---|---|
| Coding | 17.4 |
| Agentic | 2.1 |
Sources are listed separately rather than averaged. Across the 62 models both have scored they correlate at r = 0.806 — close agreement overall, but four models rank very differently between them, and a blended score would hide exactly those.
Capability
| Context window | 262K |
|---|---|
| Max output | 33K |
| Input modes | text |
| Tool use | yes |
| Reasoning | always on |
| Knowledge cutoff | 2025-09-30 |
Provenance
| Price source | openrouter.ai |
|---|---|
| Fetched | 2026-08-24 |
| Quality data | verified |
Qwen3-Next-80B-A3B-Thinking is a reasoning-first chat model in the Qwen3-Next line that outputs structured “thinking” traces by default. It’s designed for hard multi-step problems; math proofs, code synthesis/debugging, logic, and agentic...