aiintelligence

DeepSeek V4 Flashlatest from deepseek

It is an AI model designed for fast processing of large amounts of text.

deepseek’s newest large language model, able to hold about 1049k tokens of context (roughly 786k words) in one conversation, working across text->text. You reach it through an API or through tools built on it. Built for developers needing fast AI model processing. deepseek describes it as: “DeepSeek V4 Flash is an efficiency-optimized Mixture-of-Experts model from DeepSeek with 284B total parameters and 13B activated parameters, supporting a 1M-token context window.” source ↗

Our tested score
88/100 · Humanising AI text
12 tasks, marked blind by 3 judges · 16 Aug 2026 · receipts
83/100 · Book writing
12 tasks, marked blind by 3 judges · 16 Aug 2026 · receipts
83/100 · Health questions
12 tasks, marked blind by 3 judges · 16 Aug 2026 · receipts
73/100 · Legal questions
12 tasks, marked blind by 3 judges · 16 Aug 2026 · receipts
88/100 · Emotional support
12 tasks, marked blind by 3 judges · 16 Aug 2026 · receipts
68/100 · Travel planning
12 tasks, marked blind by 3 judges · 16 Aug 2026 · receipts
80/100 · Translation
12 tasks, marked blind by 3 judges · 16 Aug 2026 · receipts
83/100 · Creative writing
12 tasks, marked blind by 3 judges · 16 Aug 2026 · receipts
89/100 · Research
12 tasks, marked blind by 3 judges · 16 Aug 2026 · receipts
86/100 · CV writing
12 tasks, marked blind by 3 judges · 16 Aug 2026 · receipts
84/100 · Presentations
12 tasks, marked blind by 3 judges · 16 Aug 2026 · receipts
82/100 · Vibe coding
12 tasks, marked blind by 3 judges · 16 Aug 2026 · receipts
87/100 · Job applications
12 tasks, marked blind by 3 judges · 16 Aug 2026 · receipts
87/100 · Social posts
12 tasks, marked blind by 3 judges · 16 Aug 2026 · receipts
82/100 · Flashcards
12 tasks, marked blind by 3 judges · 16 Aug 2026 · receipts
86/100 · Coding
18 tasks, marked blind by 3 judges · 16 Aug 2026 · receipts
83/100 · Study notes
12 tasks, marked blind by 3 judges · 14 Aug 2026 · receipts
93/100 · Customer replies
12 tasks, marked blind by 3 judges · 14 Aug 2026 · receipts
93/100 · Maths
12 tasks, marked blind by 3 judges · 14 Aug 2026 · receipts
85/100 · Emails
12 tasks, marked blind by 3 judges · 14 Aug 2026 · receipts
96/100 · Extraction
12 tasks, marked blind by 3 judges · 13 Aug 2026 · receipts
83/100 · Summarising
12 tasks, marked blind by 3 judges · 13 Aug 2026 · receipts
81/100 · Essays
12 tasks, marked blind by 3 judges · 13 Aug 2026 · receipts
93/100 · Spreadsheets
12 tasks, marked blind by 3 judges · 13 Aug 2026 · receipts
85/100 · Best free
30 tasks, marked blind by 3 judges · 11 Aug 2026 · receipts
API price (per 1M in/out)
$0.0798 / $0.1596
OpenRouter API17 Aug 2026
Community Elo (text)
1431
LMArena leaderboard (CC-BY-4.0)17 Aug 2026
Battle record
0W – 1L
in our published battles
Forum sentiment (90d)
69%
developer-forum posts: 16 opinions in 85
What people are posting (90 days)
Loudest post this quarter
“I canceled Claude and coded 7 days straight with DeepSeek V4 Flash 0731 — the honest cost & quality breakdown”
Battles fought
Gemini 3.5 Flash vs DeepSeek V4 Flash · Best free9–2Gemini 3.5 Flash wins

Compare DeepSeek V4 Flash against…

GPT-5.6 Luna · spreadsheets & ExcelGPT-5.6 Terra · spreadsheets & ExcelGPT-5.3-Codex · spreadsheets & ExcelGPT-5.6 Sol · spreadsheets & ExcelGPT-5.5 · spreadsheets & ExcelGLM 5.2 · spreadsheets & ExcelQwen3.7 Max · spreadsheets & ExcelGrok 4.5 · spreadsheets & ExcelClaude Opus 4.8 · spreadsheets & ExcelGPT-5.6 Sol · customer service repliesGPT-5.5 · customer service repliesGrok 4.5 · customer service repliesGPT-5.3-Codex · customer service repliesGPT-5.6 Terra · customer service repliesGemini 3.1 Pro Preview · customer service repliesQwen3.7 Max · customer service repliesGemini 3.5 Flash · customer service repliesClaude Opus 4.6 · customer service replies
Every claim we hold (6)