DeepSeek R1 Distill Qwen 1.5B

DeepSeek

Verified against Artificial Analysis · Jul 13, 2026

Capability

percentile index

Price /M

$0 / $0

in / out

Context

—

Avg score

17.5

9 benchmarks

AA Intelligence

3.7

index

Weights

Open

Released

Jan 20, 2025

About

DeepSeek R1 Distill Qwen 1.5B is a open-weights model in the DeepSeek family from DeepSeek. Benchmarked on 9 evals, averaging 17.5.

Benchmark scores · 9

avg 17.5 — higher is better

MATHartificial-analysis · 2026-07

68.7

MMLU-Proartificial-analysis · 2026-07

26.9

AIMEartificial-analysis · 2026-07

IFBenchartificial-analysis · 2026-07

13.2

GPQA Diamondartificial-analysis · 2026-07

9.8

LiveCodeBenchartificial-analysis · 2026-07

SciCodeartificial-analysis · 2026-07

6.6

Humanity's Last Examartificial-analysis · 2026-07

3.3

AA Long-Context Reasoningartificial-analysis · 2026-07

0.3

The research behind this model

all papers mentioning it →

ThoughtFold: Folding Reasoning Chains via Introspective Preference Learning▲ 25 on HF · Jun 2, 2026 N-GRPO: Embedding-Level Neighbor Mixing for Enhanced Policy Optimization▲ 24 on HF · Jun 9, 2026 ESPO: Early-Stopping Proximal Policy Optimization▲ 19 on HF · May 28, 2026 Information-Aware KV Cache Compression for Long Reasoning▲ 11 on HF · Jun 25, 2026 RoPE-Aware Bit Allocation for KV-Cache Quantization▲ 8 on HF · Jun 23, 2026

Mentions matched by name in title/abstract — from the arXiv + HF daily corpus.

Details

PublisherDeepSeek

FamilyDeepSeek

Parameters—

License—

WeightsOpen

ReleasedJan 20, 2025

History

full ledger →

Capability over time

Append-only ledger — every observed change to this model's numbers.

API providers

Artificial Analysis$0/M