Skip to content
VECTOR WIREAI INTELLIGENCE
⌘K
UTC
Models741Deals30d261Policy30d65, 14 pendingSources1,013

DeepSeek-R1-Zero

DeepSeekWeights published Jan 20, 2025Unpriced

Score basis

DeepSeek-R1-Zero is behind the leaders in reasoning. Too few results yet to rate coding, agentic tasks, safety, long context, math, multimodal tasks, multilingual tasks, instruction following, or factuality.

Capability profile

Bars show the model's median result as a share of the leading model's, per capability. Select a row to see its results.

Limited1 capability
  1. Reasoning−45.6%1 of 6
Not rated9 capabilities

Too few results yet to rate Coding, Agentic, Safety, Long Context, Math, Multimodal, Multilingual, Instruction Following or Factuality.

Price

No current price is tracked for this model. See the rate card

Evidence

22results on21benchmarks

  • 22 vendor-reported

From 2 sources · latest Oct 7, 2026 · How verification works

Research

1 paper reference DeepSeek-R1-Zero

DeepSeek-R1-Zero benchmark results

22 results on 21 benchmarks, grouped by capability. Each bar is the result as a share of the capability leader's; every result links to its source.

Reasoning

LimitedFull reasoning ranking

45.6% behind the leader1 of 6 ranked benchmarks measured

Show 1 more reasoning resultHide 1 reasoning result

Coding

Not enough dataFull coding ranking

0 of 10 ranked benchmarks measured

More results

Benchmarks outside the capability baskets. They are not ranked against a leader.

Show 13 more resultsHide 13 results

DeepSeek-R1-Zero: common questions

Who makes DeepSeek-R1-Zero?

DeepSeek-R1-Zero is made by DeepSeek.

When was DeepSeek-R1-Zero released?

DeepSeek-R1-Zero's weights were first published on Hugging Face on Jan 20, 2025.

What is DeepSeek-R1-Zero good at?

DeepSeek-R1-Zero is behind the leaders in reasoning. Too few results yet to rate coding, agentic tasks, safety, long context, math, multimodal tasks, multilingual tasks, instruction following, or factuality.

How many benchmarks has DeepSeek-R1-Zero been tested on?

We track 22 results for DeepSeek-R1-Zero on 21 benchmarks from 2 sources. The latest was recorded on Oct 7, 2026.

About this record

Where DeepSeek-R1-Zero's numbers come from, and every name it appears under.

Tracked since
May 3, 2026
Newest source mention
May 3, 2026

Where the results come from

Verification: 22 scores · 0 independently verified · 22 vendor-reported. How these tiers are assigned

From 2 sources on 2 sites. arxiv.org supplies 19 of them. Bars are coloured by trust tier.

  • arxiv.org19
  • api.llm-stats.com3

Also known as

r1-zero