Evalry
All models

Qwen: Qwen3 Next 80B A3B Thinking

by Qwen

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 planning, and reports strong results across knowledge, reasoning, coding, alignment, and multilingual evaluations. Compared with prior Qwen3 variants, it emphasizes stability under long chains of thought and efficient scaling during inference, and it is tuned to follow complex instructions while reducing repetitive or off-task behavior. The model is suitable for agent frameworks and tool use (function calling), retrieval-heavy workflows, and standardized benchmarking where step-by-step solutions are required. It supports long, detailed completions and leverages throughput-oriented techniques (e.g., multi-token prediction) for faster generation. Note that it operates in thinking-only mode.

Avg Score

80.8%

13 answers

Avg Latency

28.8s

9 runs

Pricing

$0.15

input

/

$1.20

output

per 1M tokens

Context

262K

tokens

Alternatives

Models with similar or better quality but different tradeoffs

No alternatives found

Run benchmarks on this model to discover alternatives

Other Models from Qwen

Compare performance with other models from the same creator

ModelScoreLatencyCost/1M
Qwen: Qwen3 Coder 480B A35B100.0%1.1sFree
Qwen: Qwen3 VL 235B A22B Instruct88.6%48.0s$0.70
Qwen: Qwen3 Max87.9%31.7s$3.60
Qwen: Qwen-Plus85.8%22.0s$0.80
Qwen: Qwen3 235B A22B Thinking 250784.6%248.4s$0.35
Qwen: Qwen3 Coder Plus84.2%20.5s$3.00
Qwen: Qwen3 Coder 480B A35B83.5%19.2s$0.58
Qwen: Qwen3 Coder 480B A35B81.3%6.2s$1.01
Qwen: Qwen3 VL 235B A22B Thinking80.4%112.0s$1.98
Qwen: Qwen3 VL 32B Instruct78.8%26.3s$1.00
Qwen: Qwen Plus 072878.5%19.6s$0.80
Qwen: Qwen3 235B A22B74.6%78.3s$0.40
Qwen: Qwen3 30B A3B Instruct 250773.1%30.8s$0.20
Qwen: Qwen3 Next 80B A3B Instruct71.4%27.9s$0.59
Qwen: Qwen3 VL 8B Thinking70.8%79.4s$1.14
Qwen: Qwen-Turbo70.7%18.6s$0.13
Qwen: Qwen Plus 072870.0%49.2s$2.20
Qwen: Qwen3 235B A22B Instruct 250769.6%25.6s$0.27
Qwen: Qwen3 Coder Flash69.2%14.1s$0.90
Qwen: Qwen3 30B A3B Thinking 250765.8%45.6s$0.20
Qwen: Qwen-Max 65.8%12.7s$4.00
Qwen: Qwen2.5 VL 72B Instruct65.0%22.1s$0.38
Qwen: Qwen3 Coder 30B A3B Instruct63.6%30.8s$0.17
Qwen: Qwen3 VL 30B A3B Thinking63.1%83.6s$0.60
Qwen2.5 72B Instruct62.9%22.2s$0.26
Qwen: Qwen3 8B62.5%111.5s$0.15
Qwen: Qwen3 14B61.5%70.4s$0.14
Qwen: Qwen VL Max61.3%36.4s$2.00
Qwen: Qwen3 32B59.6%121.7s$0.16
Qwen: QwQ 32B59.2%143.6s$0.28
Qwen: Qwen VL Plus57.9%10.4s$0.42
Qwen: Qwen3 30B A3B57.7%149.8s$0.14
Qwen: Qwen3 VL 30B A3B Instruct57.5%38.3s$0.38
Qwen: Qwen3 VL 8B Instruct54.6%133.7s$0.29
Qwen2.5 Coder 32B Instruct52.7%19.7s$0.07
Qwen: Qwen2.5 VL 32B Instruct47.3%39.1s$0.14
Qwen: Qwen2.5-VL 7B Instruct31.3%20.7sFree
Qwen: Qwen2.5 7B Instruct30.8%12.6s$0.07
Qwen: Qwen2.5-VL 7B Instruct21.7%10.9s$0.20
Qwen: Qwen2.5 Coder 7B Instruct6.7%6.4s$0.06
Qwen: Qwen3 Next 80B A3B Instruct——Free
Qwen: Qwen3 4B (free)——Free

Benchmark Performance

How this model performs across different benchmarks

No benchmark data available

Run benchmarks with this model to see performance breakdown

Price vs Performance

Compare cost efficiency across all models

Current model (baseline)
Other models (relative score)
Y-axis shows score difference from shared benchmarks. X-axis uses log scale.

Score Over Time

Performance trends across all benchmark runs

Benchmark Activity

Number of benchmark runs over time

Quickstart

Get started with this model using OpenRouter

View on OpenRouter
import { OpenRouter } from "@openrouter/sdk";

const openrouter = new OpenRouter({
  apiKey: "<OPENROUTER_API_KEY>"
});

const completion = await openrouter.chat.completions.create({
  model: "qwen/qwen3-next-80b-a3b-thinking",
  messages: [
    {
      role: "user",
      content: "Hello!"
    }
  ]
});

console.log(completion.choices[0].message.content);

Get your API key at openrouter.ai/keys