vs
0:11
overall capability wins (0 ties across 11 scored). DeepSeek V4 Pro 0813 leads overall.
- Reasoning0:5
- Coding0:2
- Tool use0:1
- Composite indices0:2
- Pricing3:0
- Specs0:3
Verdict
DeepSeek V4 Pro 0813 leads coding (widest gap: +59.6 on TermBench); DeepSeek V4 Pro 0813 offers ~7.8x larger context (1M vs 128K); DeepSeek V4 Pro 0813 allows ~6.0x more max output (384K vs 64K).
- DeepSeek V4 Pro 0813 leads coding (widest gap: +59.6 on TermBench)
- DeepSeek V4 Pro 0813 offers ~7.8x larger context (1M vs 128K)
- DeepSeek V4 Pro 0813 allows ~6.0x more max output (384K vs 64K)
- DeepSeek R1 ships open weights (MIT)
- DeepSeek R1 is ~2.1x cheaper on a blended token basis than DeepSeek V4 Pro 0813
Cost for 1M in + 250K out
Illustrative chat workload at primary-provider list prices.
- DeepSeek R1
- $1.10
- DeepSeek V4 Pro 0813
- $2.31
DeepSeek R1 is about 2.1x cheaper on this workload.
Price per 1M tokens
Full comparison
Organization
DeepSeek
DeepSeek
Family
DeepSeek
DeepSeek
License
MIT
Proprietary
Open weights
Yes
No
Release
Jan 20, 2025
Aug 13, 2026
Knowledge cutoff
-
-
API / provider
DeepSeek
DeepSeek
Modalities
text → text
text → text
Specs
Context window
128K
1M
Max output
64K
384K
Parameters
671B (37B act.)
1.6T (49B act.)
Pricing
Input $/1M
$0.55
$1.32
Output $/1M
$2.19
$3.96
Blended $/1M (3∶1)
$0.96
$1.98
Speed
tok/s
45
-
TTFT (s)
1.2
-
Reasoning
MMLU-Pro
84.0
—
GPQA Diamond
70.8
92.8
Humanity's Last Exam
8.5
41.0
AIME 2025
70.0
—
MATH-500
97.3
—
Humanity's Last Exam (with tools)
—
—
AA-Omniscience Accuracy
30.5
49.1
AA-LCR v1.1
57.7
80.3
CritPt
0.6
18.0
MMMU-Pro
—
—
IFBench
39.0
—
Chartography
—
—
Chartography (With Tools)
—
—
Coding
SWE-bench Verified
49.2
—
SWE-bench Pro
—
—
SWE-bench Multilingual
—
—
LiveCodeBench
—
—
Terminal-Bench 2.1
19.1
78.7
Aider Polyglot
56.9
—
Terminal-Bench 3
—
—
BigCodeBench
—
—
SciCode
38.3
51.0
CursorBench
—
—
SWE-Rebench
—
—
NL2Repo-Bench
—
61.5
DeepSWE
—
62.7
WebDev Arena
—
—
Terminal-Bench 4.0
—
—
CursorBench 4.0
—
—
FrontierCode v1.1 (Main)
—
—
Arena
LMArena Elo
1,398
—
| Metric | DeepSeek R1 | DeepSeek V4 Pro 0813 | Delta |
|---|---|---|---|
| Identity | |||
| Organization | DeepSeek | DeepSeek | — |
| Family | DeepSeek | DeepSeek | — |
| License | MIT | Proprietary | — |
| Open weights | Yes | No | — |
| Release | Jan 20, 2025 | Aug 13, 2026 | — |
| Knowledge cutoff | - | - | — |
| API / provider | DeepSeek | DeepSeek | — |
| Modalities | text → text | text → text | — |
| Specs | |||
| Context window | 128K | 1M | B +872K |
| Max output | 64K | 384K | B +320K |
| Parameters | 671B (37B act.) | 1.6T (49B act.) | B +929 |
| Pricing | |||
| Input $/1M | $0.55 | $1.32 | A +$0.77 |
| Output $/1M | $2.19 | $3.96 | A +$1.77 |
| Blended $/1M (3∶1) | $0.96 | $1.98 | A +$1.02 |
| Speed | |||
| tok/s | 45 | - | — |
| TTFT (s) | 1.2 | - | — |
| Reasoning | |||
| MMLU-Pro | 84.0 | — | — |
| GPQA Diamond | 70.8 | 92.8 | B +22.0 pts |
| Humanity's Last Exam | 8.5 | 41.0 | B +32.5 pts |
| AIME 2025 | 70.0 | — | — |
| MATH-500 | 97.3 | — | — |
| Humanity's Last Exam (with tools) | — | — | — |
| AA-Omniscience Accuracy | 30.5 | 49.1 | B +18.6 pts |
| AA-LCR v1.1 | 57.7 | 80.3 | B +22.6 pts |
| CritPt | 0.6 | 18.0 | B +17.4 pts |
| MMMU-Pro | — | — | — |
| IFBench | 39.0 | — | — |
| Chartography | — | — | — |
| Chartography (With Tools) | — | — | — |
| Coding | |||
| SWE-bench Verified | 49.2 | — | — |
| SWE-bench Pro | — | — | — |
| SWE-bench Multilingual | — | — | — |
| LiveCodeBench | — | — | — |
| Terminal-Bench 2.1 | 19.1 | 78.7 | B +59.6 pts |
| Aider Polyglot | 56.9 | — | — |
| Terminal-Bench 3 | — | — | — |
| BigCodeBench | — | — | — |
| SciCode | 38.3 | 51.0 | B +12.7 pts |
| CursorBench | — | — | — |
| SWE-Rebench | — | — | — |
| NL2Repo-Bench | — | 61.5 | — |
| DeepSWE | — | 62.7 | — |
| WebDev Arena | — | — | — |
| Terminal-Bench 4.0 | — | — | — |
| CursorBench 4.0 | — | — | — |
| FrontierCode v1.1 (Main) | — | — | — |
| Arena | |||
| LMArena Elo | 1,398 | — | — |
DeepSeek R1
- Context
- 128K / 64K out
- Parameters
- 671B (37B act.)
- Price
- $0.55 / $2.19
- Speed
- 45 tok/s · 1.2s TTFT
- Modalities
- text → text
- License
- MIT
Original R1 reasoning model - strong math with explicit chain-of-thought; superseded by R1-0528 / V4 for coding.
DeepSeek V4 Pro 0813
- Context
- 1M / 384K out
- Parameters
- 1.6T (49B act.)
- Price
- $1.32 / $3.96 peak · $0.66 / $1.98 off-peak
- Speed
- —
- Modalities
- text → text
- License
- Proprietary
GA release of DeepSeek-V4-Pro (Aug 13, 2026) with substantially enhanced agentic coding (Terminal-Bench 2.1 87.9, NL2Repo 61.5, DeepSWE 62.7), native Responses API support, and low/high/max thinking-effort control. HLE is 42.7 without tools and 60.0 with tools. The hosted 0813 checkpoint weights are not yet published. Peak/off-peak API pricing took effect Aug 16, 2026: $1.32 in / $3.96 out peak, $0.66 in / $1.98 out off-peak (per 1M tokens).
Benchmark charts
Winner bars are emphasized. Per-benchmark deltas sit above each chart.