vs
1:8
overall capability wins (0 ties across 9 scored). DeepSeek V4 Flash 0731 leads overall.
- Reasoning0:5
- Coding0:1
- Pricing0:3
- Specs1:2
Verdict
DeepSeek V4 Flash 0731 leads coding (widest gap: +10.0 on SciCode); DeepSeek V4 Flash 0731 offers ~6.1x larger context (1M vs 164K); DeepSeek V4 Flash 0731 allows ~6.0x more max output (384K vs 64K).
- DeepSeek V4 Flash 0731 leads coding (widest gap: +10.0 on SciCode)
- DeepSeek V4 Flash 0731 offers ~6.1x larger context (1M vs 164K)
- DeepSeek V4 Flash 0731 allows ~6.0x more max output (384K vs 64K)
- DeepSeek V4 Flash 0731 is ~1.4x cheaper on a blended token basis than DeepSeek R1-0528
Cost for 1M in + 250K out
Illustrative chat workload at primary-provider list prices.
- DeepSeek R1-0528
- $1.04
- DeepSeek V4 Flash 0731
- $0.77
DeepSeek V4 Flash 0731 is about 1.3x cheaper on this workload.
Price per 1M tokens
Full comparison
Organization
DeepSeek
DeepSeek
Family
DeepSeek
DeepSeek
License
MIT
MIT
Open weights
Yes
Yes
Release
May 28, 2025
Jul 31, 2026
Knowledge cutoff
-
-
API / provider
DeepSeek
DeepSeek
Modalities
text → text
text → text
Specs
Context window
164K
1M
Max output
64K
384K
Parameters
671B (37B act.)
304B
Pricing
Input $/1M
$0.50
$0.44
Output $/1M
$2.15
$1.32
Blended $/1M (3∶1)
$0.91
$0.66
Speed
tok/s
50
-
TTFT (s)
1
-
Reasoning
MMLU-Pro
85.0
—
GPQA Diamond
81.3
90.8
Humanity's Last Exam
15.8
38.6
AIME 2025
87.5
—
MATH-500
98.3
—
Humanity's Last Exam (with tools)
—
—
AA-Omniscience Accuracy
30.5
40.4
AA-LCR v1.1
55.7
79.7
CritPt
1.4
16.6
MMMU-Pro
—
—
IFBench
39.6
—
Chartography
—
—
Chartography (With Tools)
—
—
Coding
SWE-bench Verified
57.6
—
SWE-bench Pro
—
—
SWE-bench Multilingual
—
—
LiveCodeBench
84.4
—
Terminal-Bench 2.1
—
78.7
Aider Polyglot
71.4
—
Terminal-Bench 3
—
—
BigCodeBench
—
—
SciCode
40.3
50.3
CursorBench
—
—
SWE-Rebench
22.3
—
NL2Repo-Bench
—
54.2
DeepSWE
—
54.4
WebDev Arena
—
—
Terminal-Bench 4.0
—
—
CursorBench 4.0
—
—
FrontierCode v1.1 (Main)
—
—
Arena
LMArena Elo
1,421
—
| Metric | DeepSeek R1-0528 | DeepSeek V4 Flash 0731 | Delta |
|---|---|---|---|
| Identity | |||
| Organization | DeepSeek | DeepSeek | — |
| Family | DeepSeek | DeepSeek | — |
| License | MIT | MIT | — |
| Open weights | Yes | Yes | — |
| Release | May 28, 2025 | Jul 31, 2026 | — |
| Knowledge cutoff | - | - | — |
| API / provider | DeepSeek | DeepSeek | — |
| Modalities | text → text | text → text | — |
| Specs | |||
| Context window | 164K | 1M | B +836K |
| Max output | 64K | 384K | B +320K |
| Parameters | 671B (37B act.) | 304B | A +367 |
| Pricing | |||
| Input $/1M | $0.50 | $0.44 | B +$0.06 |
| Output $/1M | $2.15 | $1.32 | B +$0.83 |
| Blended $/1M (3∶1) | $0.91 | $0.66 | B +$0.25 |
| Speed | |||
| tok/s | 50 | - | — |
| TTFT (s) | 1 | - | — |
| Reasoning | |||
| MMLU-Pro | 85.0 | — | — |
| GPQA Diamond | 81.3 | 90.8 | B +9.5 pts |
| Humanity's Last Exam | 15.8 | 38.6 | B +22.8 pts |
| AIME 2025 | 87.5 | — | — |
| MATH-500 | 98.3 | — | — |
| Humanity's Last Exam (with tools) | — | — | — |
| AA-Omniscience Accuracy | 30.5 | 40.4 | B +9.9 pts |
| AA-LCR v1.1 | 55.7 | 79.7 | B +24.0 pts |
| CritPt | 1.4 | 16.6 | B +15.2 pts |
| MMMU-Pro | — | — | — |
| IFBench | 39.6 | — | — |
| Chartography | — | — | — |
| Chartography (With Tools) | — | — | — |
| Coding | |||
| SWE-bench Verified | 57.6 | — | — |
| SWE-bench Pro | — | — | — |
| SWE-bench Multilingual | — | — | — |
| LiveCodeBench | 84.4 | — | — |
| Terminal-Bench 2.1 | — | 78.7 | — |
| Aider Polyglot | 71.4 | — | — |
| Terminal-Bench 3 | — | — | — |
| BigCodeBench | — | — | — |
| SciCode | 40.3 | 50.3 | B +10.0 pts |
| CursorBench | — | — | — |
| SWE-Rebench | 22.3 | — | — |
| NL2Repo-Bench | — | 54.2 | — |
| DeepSWE | — | 54.4 | — |
| WebDev Arena | — | — | — |
| Terminal-Bench 4.0 | — | — | — |
| CursorBench 4.0 | — | — | — |
| FrontierCode v1.1 (Main) | — | — | — |
| Arena | |||
| LMArena Elo | 1,421 | — | — |
DeepSeek R1-0528
- Context
- 164K / 64K out
- Parameters
- 671B (37B act.)
- Price
- $0.50 / $2.15
- Speed
- 50 tok/s · 1s TTFT
- Modalities
- text → text
- License
- MIT
Updated R1 reasoning checkpoint - strong math/STEM with explicit chain-of-thought.
DeepSeek V4 Flash 0731
- Context
- 1M / 384K out
- Parameters
- 304B
- Price
- $0.44 / $1.32 peak · $0.22 / $0.66 off-peak
- Speed
- —
- Modalities
- text → text
- License
- MIT
Official DeepSeek-V4-Flash release (Jul 31, 2026 public beta) superseding the preview - same base architecture plus a DSpark speculative-decoding module (304B total), MIT weights on Hugging Face, and agent benchmarks well above V4-Pro-Preview (Terminal-Bench 2.1 82.7, NL2Repo 54.2). Peak/off-peak API pricing took effect Aug 16, 2026: $0.44 in / $1.32 out peak, $0.22 in / $0.66 out off-peak (per 1M tokens).
Benchmark charts
Winner bars are emphasized. Per-benchmark deltas sit above each chart.