Skip to main content

Qwen3 Next 80B A3B Instruct

AlibabaQwenOpen WeightApache 2.0 · Commercial OK

Description

Qwen3-Next-80B-A3B-Instruct is the first in the Qwen3-Next series, featuring groundbreaking architectural innovations. It uses Hybrid Attention combining Gated DeltaNet and Gated Attention for efficient ultra-long context modeling, High-Sparsity MoE with 512 experts (10 activated + 1 shared) achieving extreme low activation ratio, and Multi-Token Prediction for improved performance and faster inference. With 80B total parameters and only 3B activated, it outperforms Qwen3-32B-Base with 10% training cost and 10x throughput for 32K+ contexts. The model performs on par with Qwen3-235B-A22B-Instruct-2507 while excelling at ultra-long-context tasks up to 256K tokens (extensible to 1M with YaRN). Architecture: 48 layers, 15T training tokens, hybrid layout of 12*(3*(Gated DeltaNet->MoE)->(Gated Attention->MoE)).

Release Date
2025-09-11
Parameters
80.0B
Context Length
131K
Modalities
text

Capability Radar

70
general
50
coding
70
reasoning
60
scienceest.
50
agents
0
multimodal

Science is estimated from LLM Stats science scores or reasoning when no dedicated science benchmarks are available.

Rankings

No ranking data available

Benchmark Scores (LLM Stats)

(LLM Stats (zeroeval))

Chat

IFEvalGoogle Research (2023)87.6%SR
Multi-IF75.8%SR
TAU-bench Retail60.9%SR
Tau2 Airline45.5%SR

Communication

Tau2 Retail57.3%SR
Tau2 Telecom13.2%SR

Creativity

Creative Writing v385.3%SR

General

MultiPL-E87.8%SR
Arena-Hard v282.7%SR
Include78.9%SR
BFCL-v370.3%SR
Aider-Polyglot49.8%SR

Language

MMLU-Redux90.9%SR
MMLU-Pro80.6%SR
MMLU-ProX76.7%SR

Math

LiveBench 2024112575.8%SR
AIME 202569.5%SR
HMMT2554.1%SR
PolyMATH45.9%SR

Reasoning

GPQANYU + Cohere + Anthropic (2023)72.9%SR
SuperGPQA58.8%SR
LiveCodeBench v656.6%SR
TAU-bench Airline44.0%SR

Writing

WritingBench87.3%SR

AA Evaluation Indices

(Artificial Analysis)

No AA evaluation data available

LLM Stats Category Scores

(LLM Stats (zeroeval))
Creativity
90
Writing
90
Instruction Following
80
Language
80
Legal
80
Structured Output
80
Chat
70
Math
70
Physics
70
Finance
70
General
70
Healthcare
70
Agents
70
Biology
70
Chemistry
70
Reasoning
60
Economics
60
Multimodal
50
Spatial Reasoning
50
Code
50
Communication
50
Tool Calling
50
Vision
50

Pricing

Input Price$0.1 / 1M tokens
Output Price$1.1 / 1M tokens
Blended Price (3:1)$0.3425 / 1M tokens

Speed

Tokens/sec153.9
Time to First Token0.83s
Time to Answer0.83s

Provider Price Ranking

Provider Price Ranking

11 providers

Cheapest: DeepInfraMost Expensive: Neon
ProviderInputOutput
1DeepInfraCheapest
$0.09
$1.1
2NanoGPT
$0.09
$1.1
3OpenRouter
$0.09
$1.1
4Kilo Gateway
$0.09
$1.1
5AlibabaPRIMARY
$0.1
$1.1
6Merge Gateway
$0.1
$1.1
7Helicone
$0.14
$1.4
8Alibaba (China)
$0.144
$0.574
9Qiniu
$0.14493
$0.57971
10Jiekou.AI
$0.15
$1.5
11Neon
$0.15
$1.2

Compare pricing across different API providers for this model.

External Sources