DeepSeek

DeepSeek V4.1 FlashOffen

Name: DeepSeek V4.1 Flash
Price: 1.2 USD
Author: DeepSeek

DeepSeek V4.1 Flash is an open-weights MoE reasoning model with 552B total (16B active) parameters, 1.0M context, and chain-of-thought problem solving.

#317veröffentlicht 10. Sept. 2026·Open Weights·552B (16B active)

textimage1.0M ctxreasoning

API providers Rangliste

Intelligence

54.5 #317

AAQI 40.0

Output speed

214.4 t/s

TTFT 1.21 s · ITL 4.7 ms/tok

Blended price (3:1)

$0.18 /1M

In $0.3/M · Out $1.2/M

Context

1.0M

≈ 2,500 A4 pages · max out 32768

Quality Domain Radar

6-Axis Frontier Evaluation vs Category Average & Flagship

DeepSeek V4.1 Flash

Claude Mythos Preview (#1)

Average (68)

Radar axes are approximated from category scores where direct benchmark measurements are missing — see Methodik.

API deployment

Compare hosts & latency

Throughput, TTFT and cache pricing across Groq, Cerebras, Together, DeepInfra, Fireworks, Bedrock and Azure.

GroqCerebrasTogether AIDeepInfraFireworks AIAWS BedrockAzure AI

View providers matrix →

Ergebnisse nach Quelle

Arena.ai ↗

1450.0

#68 · 15.4k votes · ±6

Artificial Analysis ↗

Kategorie-Punkte

Long Context98.0 Value92.0 Math68.0 Speed68.0 Reasoning58.0 Coding52.0 Multimodal50.0 Agentic48.0 Knowledge46.0 intelligence40.0

Benchmark-Ergebnisse

Mmlupro72%

LiveCodeBench68%SWE-Bench Verified65%GPQA Diamond62%

Modelldetails

Veröffentlichungsdatum

10. Sept. 2026

Parameter

552B (16B active)

Lizenz

Open Weights

Kontextfenster

1,000,000 tokens

Eingabepreis

$0.3/M

Ausgabepreis

$1.2/M

Ausgabegeschwindigkeit

214 t/s

Zeit bis zum ersten Token

1.21 s

Estimated monthly bill

Monthly API requests: 100,000Avg input tokens: 1,500Avg output tokens: 500Prompt cache hit rate: 50%

Estimated monthly spend

$0.00

Calculating savings vs GPT-4o...

Cite & embed

GitHub Markdown badge

[![LLMPodium Rank](https://llmpodium.com/badge/deepseek-v4-1-flash/rank.svg)](https://llmpodium.com/models/deepseek-v4-1-flash)

BibTeX

@misc{llmpodium2026deepseekv41flash,
  title = {Benchmark Analysis and Intelligence Evaluation of DeepSeek V4.1 Flash},
  author = {LLMPodium Research},
  year = {2026},
  url = {https://llmpodium.com/models/deepseek-v4-1-flash}
}