NOUMENOpen app
NOUMEN

DeepSeek · Open weights

DeepSeek Chat 0324

Added Jan 1, 2024Open weights

Compare models

Choose up to 3 models to compare.0 of 3 chosen

Loading models…

Use in the app

DeepSeek V3 0324, DeepSeek's 03 March 2025 V3 model, optimized for general-purpose tasks. Quantized at FP8.

  • Tools
  • Structured output
Input
Text
Output
Text
Discovery tier
Open weights

Pricing

USD per 1M tokens
Input
$0.20
Output
$0.77
Cache read
$0.14

Reference prices as published by the model provider. API example

Specifications

Context window
128K
Max output
8.2K
Average output (7 days)
280 tokens
Longer than 29% of 465 models

Side by side

Compare models

Put DeepSeek Chat 0324 next to up to three other models. The comparison stays in the link, so you can share it.

Compare models

Choose up to 3 models to compare.0 of 3 chosen

Loading models…

TryQwen3 Coder FlashDiffusionGemmaMiniMax M1All 3

Public evaluations

Benchmarks

Independent evaluation results, with how each score ranks among the models in the Noumen catalog.

Intelligence Index

9.7

Better than 19% of 326 scored models

3.8Catalog models by score55.1

Indexes

  • Coding Index

    21.2

    Composite of the coding evaluations

    Better than 15% of 168 scored models

  • Math Index

    41.0

    Composite of the math evaluations

    Better than 40% of 118 scored models

Agentic work

  • AutomationBench-AA

    0.3%

    Workflow automation, penalized for guardrail violations

    Better than 4% of 173 scored models

  • AutomationBench-AA tasks completed

    0.0%

    Workflows finished end to end without a guardrail violation

    Better than 0% of 173 scored models

  • AA-Briefcasev1.1

    68 Elo

    Agentic knowledge work, rated head to head

    Better than 0% of 166 scored models

  • GDPval-AAv2.1

    69 Elo

    Economically valuable occupational tasks, rated head to head

    Better than 0% of 187 scored models

Document reasoning

  • GDP.pdf

    1.6%

    Professional reasoning over PDF documents, all-pass rate

    Better than 7% of 171 scored models

  • AA-LCRv1.1

    40.7%

    Reasoning across long documents

    Better than 18% of 309 scored models

Reasoning and knowledge

  • Humanity’s Last Exam

    4.7%

    Expert-level questions across academic fields

    Better than 14% of 323 scored models

  • GPQA Diamond

    65.5%

    Graduate-level science questions

    Better than 22% of 291 scored models

  • MMLU-Pro

    81.9%

    Multi-subject knowledge and reasoning

    Better than 62% of 132 scored models

  • IFBench

    41.0%

    Following precise output instructions

    Better than 29% of 215 scored models

Coding

  • Terminal-Benchv4.0

    0.0%

    Practical software engineering in a terminal

    Better than 0% of 171 scored models

  • SciCode

    39.0%

    Scientific computing in Python

    Better than 19% of 174 scored models

  • LiveCodeBench

    40.5%

    Recent competitive programming problems

    Better than 34% of 129 scored models

Math

  • AIME 2025

    41.0%

    Competition mathematics, 2025 problems

    Better than 40% of 118 scored models

  • AIME

    52.0%

    Competition mathematics

    Better than 63% of 72 scored models

  • MATH-500

    94.2%

    Competition math problems

    Better than 63% of 73 scored models

Legacy evaluations

  • Terminal-Bench Hard

    15.2%

    Earlier terminal task suite

    Better than 31% of 212 scored models

  • τ²-Bench Telecom

    47.1%

    Tool use in a customer-support simulation

    Better than 40% of 214 scored models

  • AA-LCR (unversioned)

    40.7%

    Long-context reasoning, earlier grading

    Better than 18% of 309 scored models

Scores from Artificial Analysis, updated Oct 26, 2026. Artificial Analysis

Live routing data

Performance and reliability

Recent averages across requests routed for this model. Your own speed depends on prompt length, reasoning effort and load.

Throughput
27.0 tok/s
Faster than 17% of 247 models
Time to first token
2.6 s
Quicker than 63% of 247 models
Prompt cache hits
70.0%
About $0.15 per 1M input tokens with caching
Request success (24h)
99.9%
Reliable

Cache hits are the share of input tokens recently served from a prompt cache across all requests. They are not a forecast of your own hit rate.

Hourly request success · last 72 hoursLowest hour 98.2%
  • Healthy (98% or more)
  • Degraded (90–98%)
  • Down (under 90%)
  • No requests

Developers

Use it through the API

Noumen’s API is OpenAI-compatible: any Chat Completions client works with a base URL and a key.

Read the quickstart
curl https://www.noumen.xyz/api/v1/chat/completions \
  -H "Authorization: Bearer $NOUMEN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "deepseek-v3-0324",
    "messages": [{ "role": "user", "content": "Explain this stack trace." }]
  }'
import os
from openai import OpenAI

client = OpenAI(base_url="https://www.noumen.xyz/api/v1", api_key=os.environ["NOUMEN_API_KEY"])
reply = client.chat.completions.create(
    model="deepseek-v3-0324",
    messages=[{"role": "user", "content": "Explain this stack trace."}],
)
print(reply.choices[0].message.content)

DeepSeek

Other DeepSeek models, starting with the same family.