Model

Qwen-1.8B

OPEN

Alibaba · Qwen family · released Nov 30, 2023

Smallest Gen-1 model (2.2T tokens) aimed at edge/low-latency deployment; research-only license unless commercial terms are separately granted.

ReasoningCodingVisionFunction callingTool useAgentic

-157.9

Elo · rank #419

Parameters

1.8B

Active params

1.8B (dense)

Context

32K tokens

Architecture

Dense decoder-only transformer (RoPE, SwiGLU, RMSNorm), tied embeddings

License

Tongyi Qianwen RESEARCH LICENSE AGREEMENT (non-commercial by default; commercial use requires requesting a separate license from Alibaba)

Languages

—

API price (in/out)

No hosted API

Modalities

text

Benchmark results

Bar shows position within the tracked field; marker = field best

BIG-Bench HardReasoning22.3%#136

best: ERNIE 4.5 300B-A47B · 94.3%

C-EvalKnowledge56.2%#48

best: Qwen3.6-Plus · 93.3%

CMMLUKnowledge52.1%#27

best: Doubao-1.5-Pro · 90.9%

GSM8KMath32.3%#152

best: Llama 3.1 405B · 96.8%

HumanEvalCoding15.2%#177

best: Claude Opus 4.5 · 99.4%

MATH-500Math2.3%#204

best: GPT-5 · 99.4%

MBPPCoding14.2%#109

best: Llama-3.3-Nemotron-Super-49B v1 (Reasoning On) · 91.3%

MMLUKnowledge45.3%#246

best: OpenAI o3 · 92.9%

Run it locally

VRAM @ Q4

1.2 GB

VRAM @ FP16

3.6 GB

Fits on (Q4)

RTX 3060 12GBRTX 4070 Ti 16GBRTX 3090 24GBRTX 4090 24GBRTX 5090 32GBM4 Pro 48GBM3 Max 128GBM3 Ultra 512GBA100 80GBH100 80GBH200 141GBB200 192GB

Throughput data unavailable.

Quantizations

GPTQ Int4 · GPTQ Int8 · GGUF

Qwen family

Elo progression across releases

Qwen-7BAug 2023137.5 Qwen-14BSep 2023347.6 Qwen-1.8BNov 2023-157.9 Qwen-72BNov 2023535.8

Resources

HFQwen/Qwen-1_8B GHQwenLM/Qwen

API price weights · each benchmark row carries its own source badge (see methodology)