DeepSeek V4 Flash 0731
deepseek/deepseek-v4-flash-20260731
DeepSeek V4 Flash 0731 is a text-to-text model with a 1,048,576 token context window. It supports tool use and reasoning. The model is available as open weights.
- tool use
- structured outputs
- reasoning
- open-weight
At a glance
Specifications
What the API gives you
- Context window
- 1.3M tokens
- Max output
- 944K tokens
- Input price
- $0.060 / 1M tokens
- Output price
- $0.12 / 1M tokens
- Cache read
- $0.012 / 1M tokens
- Modalities
- text → text
- Knowledge cutoff
- —
- Released
- Jul 31, 2026
Measured evidence
Public benchmark record
- Design Arena Coding
- 1258 Elo2026-08-09
- Design Arena Coding
- 1259 Elo2026-08-14
- Design Arena Coding
- 1268 Elo2026-08-06
- Design Arena Coding
- 1257 Elo2026-08-10
- Design Arena Coding
- 1260 Elo2026-08-07
- Design Arena Coding
- 1253 Elo2026-08-18
- Design Arena Coding
- 1258 Elo2026-08-08
- Design Arena Coding
- 1253 Elo2026-08-15
- Design Arena Coding
- 1260 Elo2026-08-11
- Design Arena Coding
- 1260 Elo2026-08-12
- Design Arena Coding
- 1259 Elo2026-08-13
- Design Arena Coding
- 1253 Elo2026-08-19
- Design Arena Coding
- 1256 Elo2026-08-27
- Design Arena Coding
- 1260 Elo2026-08-21
- Design Arena Coding
- 1253 Elo2026-08-20
- Design Arena Coding
- 1253 Elo2026-08-16
- Design Arena Coding
- 1248 Elo2026-09-05
- Design Arena Coding
- 1254 Elo2026-08-17
- Design Arena Coding
- 1259 Elo2026-08-22
- Design Arena Coding
- 1259 Elo2026-08-23
- Design Arena Coding
- 1256 Elo2026-08-28
- Design Arena Coding
- 1258 Elo2026-08-24
- Design Arena Coding
- 1257 Elo2026-08-25
- Design Arena Coding
- 1257 Elo2026-08-26
- Design Arena Coding
- 1248 Elo2026-09-06
- Design Arena Coding
- 1255 Elo2026-08-29
- Design Arena Coding
- 1255 Elo2026-08-30
- Design Arena Coding
- 1248 Elo2026-09-07
- Design Arena Coding
- 1255 Elo2026-08-31
- Design Arena Coding
- 1255 Elo2026-09-01
- Design Arena Coding
- 1254 Elo2026-09-02
- Design Arena Coding
- 1249 Elo2026-09-08
- Design Arena Coding
- 1251 Elo2026-09-03
- Design Arena Coding
- 1251 Elo2026-09-04
- Design Arena Coding
- 1247 Elo2026-09-09
- Design Arena Coding
- 1246 Elo2026-09-10
- Design Arena Coding
- 1246 Elo2026-09-11
- Design Arena Coding
- 1245 Elo2026-09-12
- Design Arena Coding
- 1245 Elo2026-09-13
- Design Arena Coding
- 1245 Elo2026-09-14
- Design Arena Coding
- 1245 Elo2026-09-15
- Design Arena UI
- 1273 Elo2026-08-09
- Design Arena UI
- 1265 Elo2026-08-14
- Design Arena UI
- 1259 Elo2026-08-06
- Design Arena UI
- 1252 Elo2026-08-07
- Design Arena UI
- 1272 Elo2026-08-10
- Design Arena UI
- 1268 Elo2026-08-08
- Design Arena UI
- 1269 Elo2026-08-18
- Design Arena UI
- 1273 Elo2026-08-11
- Design Arena UI
- 1264 Elo2026-08-15
- Design Arena UI
- 1276 Elo2026-08-12
- Design Arena UI
- 1263 Elo2026-08-13
- Design Arena UI
- 1269 Elo2026-08-19
- Design Arena UI
- 1263 Elo2026-08-16
- Design Arena UI
- 1269 Elo2026-08-20
- Design Arena UI
- 1265 Elo2026-08-27
- Design Arena UI
- 1270 Elo2026-08-17
- Design Arena UI
- 1261 Elo2026-09-05
- Design Arena UI
- 1266 Elo2026-08-21
- Design Arena UI
- 1267 Elo2026-08-22
- Design Arena UI
- 1266 Elo2026-08-23
- Design Arena UI
- 1264 Elo2026-08-28
- Design Arena UI
- 1268 Elo2026-08-24
- Design Arena UI
- 1268 Elo2026-08-25
- Design Arena UI
- 1268 Elo2026-08-26
- Design Arena UI
- 1262 Elo2026-09-06
- Design Arena UI
- 1264 Elo2026-08-29
- Design Arena UI
- 1266 Elo2026-08-30
- Design Arena UI
- 1258 Elo2026-09-07
- Design Arena UI
- 1266 Elo2026-08-31
- Design Arena UI
- 1267 Elo2026-09-01
- Design Arena UI
- 1262 Elo2026-09-02
- Design Arena UI
- 1259 Elo2026-09-03
- Design Arena UI
- 1255 Elo2026-09-08
- Design Arena UI
- 1260 Elo2026-09-04
- Design Arena UI
- 1255 Elo2026-09-09
- Design Arena UI
- 1251 Elo2026-09-10
- Design Arena UI
- 1250 Elo2026-09-11
- Design Arena UI
- 1252 Elo2026-09-12
- Design Arena UI
- 1252 Elo2026-09-13
- Design Arena UI
- 1252 Elo2026-09-14
- Design Arena UI
- 1252 Elo2026-09-15
- Design Arena Website
- 1259 Elo2026-08-09
- Design Arena Website
- 1260 Elo2026-08-14
- Design Arena Website
- 1269 Elo2026-08-06
- Design Arena Website
- 1263 Elo2026-08-07
- Design Arena Website
- 1259 Elo2026-08-10
- Design Arena Website
- 1259 Elo2026-08-08
- Design Arena Website
- 1252 Elo2026-08-18
- Design Arena Website
- 1260 Elo2026-08-11
- Design Arena Website
- 1252 Elo2026-08-15
- Design Arena Website
- 1261 Elo2026-08-12
- Design Arena Website
- 1261 Elo2026-08-13
- Design Arena Website
- 1252 Elo2026-08-19
- Design Arena Website
- 1252 Elo2026-08-16
- Design Arena Website
- 1252 Elo2026-08-20
- Design Arena Website
- 1260 Elo2026-08-27
- Design Arena Website
- 1252 Elo2026-08-17
- Design Arena Website
- 1255 Elo2026-09-05
- Design Arena Website
- 1264 Elo2026-08-21
- Design Arena Website
- 1264 Elo2026-08-22
- Design Arena Website
- 1263 Elo2026-08-23
- Design Arena Website
- 1262 Elo2026-08-24
- Design Arena Website
- 1260 Elo2026-08-28
- Design Arena Website
- 1261 Elo2026-08-25
- Design Arena Website
- 1261 Elo2026-08-26
- Design Arena Website
- 1252 Elo2026-09-06
- Design Arena Website
- 1259 Elo2026-08-29
- Design Arena Website
- 1259 Elo2026-08-30
- Design Arena Website
- 1259 Elo2026-08-31
- Design Arena Website
- 1251 Elo2026-09-07
- Design Arena Website
- 1259 Elo2026-09-01
- Design Arena Website
- 1259 Elo2026-09-02
- Design Arena Website
- 1256 Elo2026-09-03
- Design Arena Website
- 1252 Elo2026-09-08
- Design Arena Website
- 1255 Elo2026-09-04
- Design Arena Website
- 1251 Elo2026-09-09
- Design Arena Website
- 1251 Elo2026-09-10
- Design Arena Website
- 1251 Elo2026-09-11
- Design Arena Website
- 1250 Elo2026-09-12
- Design Arena Website
- 1250 Elo2026-09-13
- Design Arena Website
- 1250 Elo2026-09-14
- Design Arena Website
- 1250 Elo2026-09-15
- LiveBench Agentic Coding
- 46.8%2026-06-25
- LiveBench Coding
- 75.0%2026-06-25
- LiveBench Data Analysis
- 79.3%2026-06-25
- LiveBench Instruction Following
- 65.5%2026-06-25
- LiveBench Language
- 79.2%2026-06-25
- LiveBench Math
- 86.8%2026-06-25
- LiveBench Reasoning
- 86.6%2026-06-25
LMArena data licensed CC-BY-4.0. Aider (Apache-2.0) and SWE-bench (MIT) are open. Design Arena uses blind human preference.
Open-weight model on Hugging Face.
Task evidence
How broad is the support?
Each status is exact to this model version and the selected task. Official release claims remain provisional.
- Best LLM for Frontend & UIEstablished evidence
70% of intended weighted task evidence across 2 independent sources. 5 of 5 configured evaluation feeds currently have results.
- Best LLM for Competition MathProvisional evidence
65% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LMArena Instruction Following or LMArena Math. 5 of 5 configured evaluation feeds currently have results.
- Best LLM for Math & ReasoningProvisional evidence
65% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LMArena Instruction Following or LMArena Math. 5 of 5 configured evaluation feeds currently have results.
- Best LLM for Logic & Problem SolvingProvisional evidence
60% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LMArena Instruction Following or LMArena Math. 5 of 5 configured evaluation feeds currently have results.
- Best LLM for a General AssistantProvisional evidence
55% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LMArena Instruction Following or LMArena Text. 5 of 5 configured evaluation feeds currently have results.
- Best LLM for Code CompletionProvisional evidence
50% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Aider Polyglot or LMArena WebDev. 4 of 4 configured evaluation feeds currently have results.
- Best LLM for Creative WritingProvisional evidence
50% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LMArena Creative Writing or LMArena Text. 5 of 5 configured evaluation feeds currently have results.
- Best LLM for WritingProvisional evidence
50% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LMArena Creative Writing or LMArena Text. 5 of 5 configured evaluation feeds currently have results.
- Best LLM for Fast SummariesProvisional evidence
50% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LMArena Text or Route reliability. 3 of 4 configured evaluation feeds currently have results.
- Best Value LLMProvisional evidence
43% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LMArena Text or Route reliability. 3 of 4 configured evaluation feeds currently have results.
- Best Cheap LLM for High VolumeProvisional evidence
42% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LMArena Text or Route reliability. 3 of 4 configured evaluation feeds currently have results.
- Cheapest Capable LLMProvisional evidence
42% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LMArena Text or Route reliability. 3 of 4 configured evaluation feeds currently have results.
- Best Open-Weight LLM for Self-HostingProvisional evidence
42% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LMArena Text or Route reliability. 3 of 4 configured evaluation feeds currently have results.
- Best LLM for Literary TranslationProvisional evidence
40% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LMArena Text or WMT translation. 4 of 5 configured evaluation feeds currently have results.
- Best LLM for Everyday TranslationProvisional evidence
40% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LMArena Text or WMT translation. 4 of 5 configured evaluation feeds currently have results.
- Best LLM for TranslationProvisional evidence
40% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LMArena Text or WMT translation. 4 of 5 configured evaluation feeds currently have results.
- Best LLM for CodingProvisional evidence
40% of intended weighted task evidence across 2 independent sources; broader independent confirmation is still missing from Aider Polyglot or LMArena WebDev. 8 of 8 configured evaluation feeds currently have results.
- Best LLM for Business WritingProvisional evidence
40% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LMArena Creative Writing or LMArena Instruction Following. 5 of 5 configured evaluation feeds currently have results.
- Best LLM for Marketing CopyProvisional evidence
40% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LMArena Creative Writing or LMArena Instruction Following. 5 of 5 configured evaluation feeds currently have results.
- Best LLM for Chat & RoleplayProvisional evidence
40% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LMArena Creative Writing or LMArena Text. 5 of 5 configured evaluation feeds currently have results.
- Best LLM for Roleplay & CharacterProvisional evidence
40% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LMArena Creative Writing or LMArena Text. 5 of 5 configured evaluation feeds currently have results.
- Best LLM for High-Volume ExtractionProvisional evidence
31% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Route reliability or Structured-output eval. 2 of 4 configured evaluation feeds currently have results.
- Best LLM for Document SummarizationProvisional evidence
30% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from FACTS Grounding or LMArena Document. 4 of 6 configured evaluation feeds currently have results.
- Best LLM for SummarizationProvisional evidence
30% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from FACTS Grounding or LMArena Document. 4 of 6 configured evaluation feeds currently have results.
- Best Local LLM for CodingProvisional evidence
30% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Aider Polyglot or LMArena WebDev. 5 of 5 configured evaluation feeds currently have results.
- Best LLM for JSON & Schema OutputProvisional evidence
28% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Berkeley Function Calling or LMArena Text. 4 of 5 configured evaluation feeds currently have results.
- Best LLM for Structured Data ExtractionProvisional evidence
28% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Berkeley Function Calling or LMArena Text. 4 of 5 configured evaluation feeds currently have results.
- Best LLM for Debugging & Code ReviewProvisional evidence
25% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Aider Polyglot or LMArena WebDev. 6 of 6 configured evaluation feeds currently have results.
- Best LLM for Image UnderstandingProvisional evidence
20% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LMArena Vision or VLMEvalKit tasks. 3 of 4 configured evaluation feeds currently have results.
- Best LLM for Agentic CodingProvisional evidence
19% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Design Arena Full-stack or Design Arena Web-apps. 8 of 8 configured evaluation feeds currently have results.
- Best LLM for Autonomous AgentsProvisional evidence
17% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Frontier-Bench or LMArena Agent. 7 of 7 configured evaluation feeds currently have results.
- Best LLM for AI AgentsProvisional evidence
15% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Berkeley Function Calling or LMArena Agent. 7 of 7 configured evaluation feeds currently have results.
- Best LLM for Tool & Function CallingProvisional evidence
10% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from Berkeley Function Calling or LMArena Agent. 4 of 5 configured evaluation feeds currently have results.
- Best LLM for Computer & Browser UseProvisional evidence
10% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from Berkeley Function Calling or LMArena Agent. 4 of 4 configured evaluation feeds currently have results.
- Best LLM for OCR & DocumentsProvisional evidence
10% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from LMArena Document or LMArena Vision. 3 of 5 configured evaluation feeds currently have results.
- Best LLM for Vision & DocumentsProvisional evidence
10% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from LMArena Document or LMArena Vision. 3 of 5 configured evaluation feeds currently have results.
- Best LLM for Grounded RAGProvisional evidence
10% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from CRAG or FACTS Grounding. 2 of 5 configured evaluation feeds currently have results.
- Best LLM for Charts & DiagramsProvisional evidence
10% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from LMArena Document or LMArena Vision. 3 of 5 configured evaluation feeds currently have results.
- Best LLM for Document ParsingProvisional evidence
10% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from LMArena Document or LMArena Vision. 3 of 5 configured evaluation feeds currently have results.
- Best LLM for Massive ContextProvisional evidence
6% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from LMArena Document or LMArena Long Query. 3 of 4 configured evaluation feeds currently have results.
- Best LLM for Hermes AgentProvisional evidence
6% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from Berkeley Function Calling or LMArena Agent. 5 of 5 configured evaluation feeds currently have results.
- Best LLM for OpenHandsProvisional evidence
6% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from Berkeley Function Calling or LMArena Agent. 5 of 5 configured evaluation feeds currently have results.
- Best LLM for GooseProvisional evidence
6% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from Berkeley Function Calling or LMArena Agent. 5 of 5 configured evaluation feeds currently have results.
- Best LLM for RAG & Long ContextProvisional evidence
6% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from FACTS Grounding or LMArena Document. 4 of 6 configured evaluation feeds currently have results.
- Best LLM for OpenClawProvisional evidence
6% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from Berkeley Function Calling or LMArena Agent. 5 of 5 configured evaluation feeds currently have results.
Recommendation coverage
Where this model appears
Not currently included in a published recommendation. The task evidence panel above says exactly what is confirmed and what is still missing.
Catalog data via OpenRouter and models.dev. Last checked within the past 6 hours.