Qwen · Model identity
Qwen3 Next 80B A3B Instruct
Qwen3-Next-80B-A3B-Instruct is an instruction-tuned chat model in the Qwen3-Next series optimized for fast, stable responses without “thinking” traces. It targets complex tasks across reasoning, code generation, knowledge QA, and multilingual...
1pricing route
What published route metadata supports
CachingHugging FaceStructured outputTool use
Input modalities
Text
Supported API parameters
frequency_penaltylogit_biaslogprobsmax_tokensmin_ppresence_penaltyrepetition_penaltyresponse_formatseedstopstructured_outputstemperaturetool_choicetoolstop_ktop_logprobstop_p
Only source-backed fields are filled
Release dateNot independently verified
Knowledge cutoffNot independently verified
Open weightsNot independently verified
OpenRouter catalog date2025-09-11
LLMPrice does not infer a release date, knowledge cutoff, or open-weight status from a model name, Hugging Face link, or pricing route.
Could this workload cost less?Compare Qwen3 Next 80B A3B Instruct with cheaper models using the same workload.
Find cheaper alternatives →| Endpoint / route | Model ID | Input | Cached | Output | Context | Source |
|---|---|---|---|---|---|---|
| OpenRouterStandard | qwen/qwen3-next-80b-a3b-instruct | $0.1 | $0.07 | $1.1 | 262K | OpenRouter API ↗ |
Hugging Face references
Presence of a Hugging Face ID is not treated as proof that model weights are open.
2 verified price changes since Aug 26, 2026
0 route availability changes recorded. No earlier prices are inferred.
Input / 1M$0.1Aug 26, 2026 → Sep 1, 2026
- Sep 1, 2026OpenRouter · Standard
Input $0.09 → $0.1 · Cached input — → $0.07
- Aug 30, 2026OpenRouter · Standard
Input $0.1 → $0.09 · Cached input $0.07 → —