Skip to content

DeepSeek V3

DeepSeekOpen weightsNon-reasoningVerifiedReleased Dec 26, 2024Input: text
Compare with Ornith-1.5-9BSource scores ↗OpenRouter ↗
Overall score
31.9
Rank 175 of 222 · 15.5–48.3
Context window
128K
128,000 tokens
Price per 1M tokens
$0.27 in · $1.10 out
Source: BenchLM
Blended price
$0.48
3:1 input:output tokens
Output speed
—
Not measured yet

Category scores

Scores per capability area with the model's rank among ranked models in that category.

  1. 1Knowledge40.2
  2. 2Instruction following#11739.6
  3. 3Reasoning25.5
Higher is better. #N is the model's rank in that category, where it is eligible to be ranked.

Benchmark results (10)

As reported in system cards, papers and public evaluations. Position compares this model with every other model reporting the same benchmark.

Knowledge

GPQA59.1#93 of 117
MMLU-Pro75.9#43 of 67
SimpleQA24.9#7 of 10

Coding

SWE-bench Verified42.0#93 of 98
LiveCodeBench37.6#10 of 10

Math

MATH-50090.2#15 of 20
AIME 202439.2#16 of 19

Reasoning

LongBench v248.7#16 of 17

Multimodal

Design Arena Website1120#79 of 91

Instruction following

IFEval86.1#30 of 53

Pricing by source

USD per 1M tokens.

SourceInputOutputCached in
List price (BenchLM)$0.27$1.10$0.070
OpenRouter deepseek/deepseek-chat$0.26$1.03—

DeepSeek's official DeepSeek-V3 launch page lists $0.27 input (cache miss) / $1.10 output per million tokens, with cache-hit input at $0.07.

Arena ratings

Human-preference ratings from LMArena. Best-rated variant shown.

Text1333#224 · 21.8K votes

Open weights

From the Hugging Face repository.

Parameters
684B
Licence
—
Downloads (30 days)
1.4M

Speed

Measured by us through OpenRouter (median of trials).

Not measured yet.

Related models

Same family

Similar score