Capability
60.9/100
field median 57.5
#48 of 217 ranked models
DeepSeek V4 Pro 0813
Released Aug 13, 2026 — see all recent releases
Decision reading
DeepSeek V4 Pro 0813 scores 60.9 out of 100 and ranks #48 of 217. This profile shows 34 source-displayable benchmark rows; its strongest eligible category is Knowledge at #17. API pricing is $0.435 input and $0.87 output per million tokens, with cached input at $0.003625.
Data as of August 13, 2026 · How the score is built
Strongest published evidence
Knowledge ranks #17. Particularly effective for knowledge-intensive tasks like research, analysis, and factual Q&A.
Validate before choosing
34 published rows leave some tracked benchmark slots empty. Coding is its lowest eligible category at #40.
Decision snapshot
Each value carries a field reference instead of floating alone. Markers compare this model with the current ranked and priced catalog; they are not absolute quality thresholds.
Price
$0.43input / $0.87 output
input median $1
cached $0.004 · blended $0.65
Speed
83tok/s
field median 93 tok/s
First token 25.67 s
Context
1Mtokens
field median 200,000
Maximum output length is tracked separately
Capability shape
Each axis shows percentile within that category’s eligible cohort. The comparison outline is the median of the six nearest public-score peers; a collapsed vertex means the category is not rank-eligible.
DeepSeek V4 Pro 0813 category percentile values
- Agentic80th percentile
- Coding71st percentile
- ReasoningNot eligible
- Knowledge70th percentile
- MathNot eligible
- MultilingualNot eligible
- MultimodalNot eligible
- Instruction followingNot eligible
The dashed outline is median of 6 nearest peers.
Eligible category ranks
- Agentic#26/128
- Coding#40/134
- ReasoningNot ranked
- Knowledge#17/55
- MathNot ranked
- MultilingualNot ranked
- MultimodalNot ranked
- Inst. FollowingNot ranked
Top decileTop quartileMid-fieldNot eligible
What it costs to get this score
Published API price against the public score. The x-axis uses a log scale; the dashed path marks models that are not beaten by a cheaper, higher-scoring option. Price uses average of published input and output rates.
The chart opens on the current model. Scroll horizontally to inspect the full price axis.
Current modelDeepSeek V4 Pro 0813 · 60.9 score · $0.65 blended per million tokens
Horizontal: blended price per million tokens, log scale · Vertical: public score
How much of this is verified
Coverage is split by category so a strong number never hides a thin evidence base. Verified means the row is tied to a published source; provisional rows remain visible but separate.
- Agentic10/10 verified
- Coding12/12 verified
- Reasoning2/2 verified
- Knowledge6/6 verified
- Math4/4 verified
- MultilingualNot measured
- MultimodalNot measured
- Inst. FollowingNot measured
Verified sourceProvisionalNot measured
Spec sheet
Each documented value carries its source. Missing fields stay visible as not sourced or not published, rather than disappearing from the page.
- API model ID
- deepseek-v4-proDeepSeek models and pricing
- Context window
- 1MDeepSeek models and pricing
- Maximum output
- Not sourced yet
- Knowledge cutoff
- Not sourced yet
- Input modalities
- textDeepSeek GitHub Copilot integration
- Output modalities
- textDeepSeek GitHub Copilot integration
- Parameters
- Not disclosed by the provider
- Availability
- DeepSeek APIDeepSeek models and pricing
- Cloud regions
- Not tracked yet
- Lifecycle
- activeDeepSeek models and pricing
- API capabilities
- Tool calling, structured outputs, and batch support are not tracked yet
- Prompt caching
- Published at $0.004 per million cached input tokens
- Self-host
- Weights are not published
- Rate limits
- Not tracked yet
Category score record
Scores and ranks appear only where published evidence can be displayed. The table keeps the score, weight, cohort, and evidence state together.
| Category | Score | Rank | Percentile | Weight | Benchmarks | Evidence |
|---|---|---|---|---|---|---|
| AgenticRank #26 of 128Percentile 80thWeight 22%10 benchmarksVerified | 57.0 | #26 of 128 | 80th | 22% | 10 benchmarks | Verified |
| CodingRank #40 of 134Percentile 71stWeight 20%12 benchmarksVerified | 54.4 | #40 of 134 | 71st | 20% | 12 benchmarks | Verified |
| ReasoningWeight 17%2 benchmarksVerified | Score pending | Not ranked | Not available | 17% | 2 benchmarks | Verified |
| KnowledgeRank #17 of 55Percentile 70thWeight 12%6 benchmarksVerified | 77.0 | #17 of 55 | 70th | 12% | 6 benchmarks | Verified |
| MathRank Not rankedWeight 5%4 benchmarksVerified | 81.8 | Not ranked | Not available | 5% | 4 benchmarks | Verified |
| MultilingualWeight 7%0 benchmarksNot measured | Not measured | Not ranked | Not available | 7% | 0 benchmarks | Not measured |
| MultimodalWeight 12%0 benchmarksNot measured | Not measured | Not ranked | Not available | 12% | 0 benchmarks | Not measured |
| Inst. FollowingWeight 5%0 benchmarksNot measured | Not measured | Not ranked | Not available | 5% | 0 benchmarks | Not measured |
Benchmark ledger
Coding opens by default. The marker compares each value with the best source-verified result in the catalog; provisional leaders do not set the reference. Expand the remaining categories for every published row.
Coding12 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| SWE-bench VerifiedSoftware Engineering Benchmark Verified | Score80.6% | Versus best verified row Best verified: Claude Opus 5 · 96% | Gap15.4 behind | WeightWeighted 16% | Provider exact |
| SWE-bench Pro | Score55.4% | Versus best verified row Best verified: Claude Mythos 5 · 80.3% | Gap24.9 behind | WeightWeighted 10% | Provider exact |
| LiveCodeBench Pass@1-COTLiveCodeBench Pass@1 with Chain-of-Thought | Score93.5% | Versus best verified row Best verified: DeepSeek V4 Pro 0813 · 93.5% | GapBest verified | WeightDisplay only | Provider exact |
| CodeforcesCodeforces Rating | Score3206.0 | Versus best verified row Best verified: DeepSeek V4 Pro 0813 · 3206.0 | GapBest verified | WeightDisplay only | Provider exact |
| SWE Multilingual | Score76.2% | Versus best verified row Best verified: Claude Opus 5 · 89.5% | Gap13.3 behind | WeightDisplay only | Provider exact |
| Terminal-Bench 2.0 | Score67.9% | Versus best verified row Best verified: GPT-5.6 Sol · 91.9% | Gap24 behind | WeightDisplay only | Provider exact |
| Vibe Code BenchVibe Code Bench v1.1 | Score49.93% | Versus best verified row Best verified: Claude Opus 4.7 · 71.00% | Gap21.1 behind | WeightDisplay only | Benchmark exact |
| Terminal-Bench 2.1Terminal-Bench 2.1 (provider run) | Score87.9% | Versus best verified row Best verified: DeepSeek V4 Pro 0813 · 87.9% | GapBest verified | WeightDisplay only | |
| NL2Repo | Score61.5% | Versus best verified row Best verified: DeepSeek V4 Pro 0813 · 61.5% | GapBest verified | WeightDisplay only | |
| deepSwe | Score62.7% | Versus best verified row Best verified: GPT-5.6 Sol · 72.7% | Gap10 behind | WeightDisplay only | |
| DSBench-FullStackDeepSeek DSBench FullStack | Score71.1% | Versus best verified row Best verified: DeepSeek V4 Pro 0813 · 71.1% | GapBest verified | WeightDisplay only | |
| DSBench-HardDeepSeek DSBench Hard | Score67.2% | Versus best verified row Best verified: DeepSeek V4 Pro 0813 · 67.2% | GapBest verified | WeightDisplay only |
Agentic10 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| Terminal-Bench 2.0 | Score67.9% | Versus best verified row Best verified: GPT-5.6 Sol · 91.9% | Gap24 behind | WeightWeighted 38% | Provider exact |
| BrowseComp | Score83.4% | Versus best verified row Best verified: GPT-5.6 Sol · 92.2% | Gap8.8 behind | WeightWeighted 28% | Provider exact |
| Terminal-Bench 2.1Terminal-Bench 2.1 (provider run) | Score87.9% | Versus best verified row Best verified: DeepSeek V4 Pro 0813 · 87.9% | GapBest verified | WeightDisplay only | |
| HLE w/ toolsHumanity's Last Exam with tools | Score60.0% | Versus best verified row Best verified: Claude Opus 5 · 64.7% | Gap4.7 behind | WeightDisplay only | |
| MCP Atlas | Score73.6% | Versus best verified row Best verified: Muse Spark 1.1 · 88.1% | Gap14.5 behind | WeightDisplay only | Provider exact |
| Toolathlon | Score51.8% | Versus best verified row Best verified: Muse Spark 1.1 · 75.6% | Gap23.8 behind | WeightDisplay only | Provider exact |
| CyberGym | Score83.3% | Versus best verified row Best verified: Fugu Cyber · 86.9% | Gap3.6 behind | WeightDisplay only | |
| Toolathlon-Verified | Score74.1% | Versus best verified row Best verified: Claude Opus 5 · 80.6% | Gap6.5 behind | WeightDisplay only | |
| Agents' Last Exam | Score25.7% | Versus best verified row Best verified: Qwen3.8 Max · 52.4% | Gap26.7 behind | WeightDisplay only | |
| AutomationBench | Score31.8% | Versus best verified row Best verified: DeepSeek V4 Pro 0813 · 31.8% | GapBest verified | WeightDisplay only |
Reasoning2 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| MRCR 1M | Score83.5% | Versus best verified row Best verified: DeepSeek V4 Pro 0813 · 83.5% | GapBest verified | WeightDisplay only | Provider exact |
| CorpusQA 1M | Score62.0% | Versus best verified row Best verified: DeepSeek V4 Pro 0813 · 62.0% | GapBest verified | WeightDisplay only | Provider exact |
Knowledge6 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| HLEHumanity's Last Exam | Score42.7% | Versus best verified row Best verified: Claude Opus 5 · 64.7% | Gap22 behind | WeightWeighted 45% | |
| MMLU-ProMassive Multitask Language Understanding Professional | Score87.5% | Versus best verified row Best verified: Qwen3.7 Max · 89.6% | Gap2.1 behind | WeightWeighted 30% | Provider exact |
| SimpleQAMeasuring Short-Form Factuality in Large Language Models | Score57.9% | Versus best verified row Best verified: DeepSeek V4 Pro 0813 · 57.9% | GapBest verified | WeightWeighted 11% | Provider exact |
| GPQAGraduate-Level Google-Proof Q&A | Score90.1% | Versus best verified row Best verified: Sakana Fugu-Ultra · 95.5% | Gap5.4 behind | WeightWeighted 7% | Provider exact |
| Chinese-SimpleQA | Score84.4% | Versus best verified row Best verified: DeepSeek V4 Pro 0813 · 84.4% | GapBest verified | WeightDisplay only | Provider exact |
| GPQA-DGPQA Diamond | Score90.1% | Versus best verified row Best verified: Sakana Fugu-Ultra · 95.5% | Gap5.4 behind | WeightDisplay only | Provider exact |
Math4 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| HMMT Feb 2026Harvard-MIT Mathematics Tournament February 2026 | Score95.2% | Versus best verified row Best verified: Qwen3.7 Max · 97.1% | Gap1.9 behind | WeightWeighted 25% | Provider exact |
| IMOAnswerBench | Score89.8% | Versus best verified row Best verified: Qwen3.7 Max · 90.0% | Gap0.2 behind | WeightDisplay only | Provider exact |
| Apex | Score38.3% | Versus best verified row Best verified: Qwen3.7 Max · 44.5% | Gap6.2 behind | WeightDisplay only | Provider exact |
| Apex Shortlist | Score90.2% | Versus best verified row Best verified: DeepSeek V4 Pro 0813 · 90.2% | GapBest verified | WeightDisplay only | Provider exact |
Lineage
The sequence follows explicit supersedes links. Scores and prices remain blank when the corresponding public row or first-party rate is unavailable.
Aug 13, 2026 · you are here
DeepSeek V4 Pro 0813Score 60.9 · $0.435 / $0.87
pro-reasoning · 0813
How to read this profile
The visual layer above carries the decisions. These notes preserve the model, ranking, coverage, and family context behind the numbers.
DeepSeek V4 Pro 0813 ranks #48 of 217 on the public leaderboard with a score of 60.93/100. It does not yet have enough sourced coverage for a verified position.
DeepSeek V4 Pro 0813 is a proprietary model with a 1M context window. It uses an explicit reasoning mode, which can improve complex problem solving while adding latency and token use.
Rolling profile for the current DeepSeek-V4-Pro-0813 API. It retains the earlier DeepSeek V4 Pro Max snapshot and overlays DeepSeek's newer August 13 max-effort results where the provider published a stronger value. HLE rises from 37.7 to 42.7 and HLE with tools rises from 48.2 to 60.0; the newly published agent and coding benchmarks are added without removing prior coverage. DeepSeek has not published weights for the 0813 hosted checkpoint.
Its explicit predecessor is DeepSeek V4 Pro 0813. 34 of 414 tracked benchmark slots currently have displayable evidence. Missing categories stay blank.
Its strongest eligible category is Knowledge at #17, while its lowest eligible position is Coding at #40. particularly effective for knowledge-intensive tasks like research, analysis, and factual Q&A.
Radar
Full release historyDeepSeek V4 Pro 0813 release history
DeepSeek V4 Pro 0813 released
DeepSeek · Model release
Source confirmed
Frequently asked questions
How does DeepSeek V4 Pro 0813 perform overall in AI benchmarks?
DeepSeek V4 Pro 0813 ranks #48 out of 217 models on the public BenchAlign leaderboard, with a score of 60.93/100. Its evidence status is Estimated, and this profile shows 34 source-displayable benchmark rows. The label describes evidence depth, not a provider quality claim; inspect category rows before choosing a workload.
Is DeepSeek V4 Pro 0813 good for knowledge and understanding?
DeepSeek V4 Pro 0813 ranks #17 out of 55 eligible models for knowledge and understanding, with a public category score of 77/100. Higher-ranked alternatives are available for workloads where this category decides the choice. Check the underlying rows before treating the aggregate as a workload guarantee.
Is DeepSeek V4 Pro 0813 good for coding and programming?
DeepSeek V4 Pro 0813 ranks #40 out of 134 eligible models for coding and programming, with a public category score of 54.4/100. Higher-ranked alternatives are available for workloads where this category decides the choice. Check the underlying rows before treating the aggregate as a workload guarantee.
Is DeepSeek V4 Pro 0813 good for mathematics?
DeepSeek V4 Pro 0813 has source-displayable benchmark coverage for mathematics, but the public category table does not assign it a rank there. The individual rows remain available for inspection. A missing category position means the evidence threshold was not met; it does not convert the model's unmeasured work into a zero.
Is DeepSeek V4 Pro 0813 good for reasoning and logic?
DeepSeek V4 Pro 0813 has source-displayable benchmark coverage for reasoning and logic, but the public category table does not assign it a rank there. The individual rows remain available for inspection. A missing category position means the evidence threshold was not met; it does not convert the model's unmeasured work into a zero.
Is DeepSeek V4 Pro 0813 good for agentic tool use and computer tasks?
DeepSeek V4 Pro 0813 ranks #26 out of 128 eligible models for agentic tool use and computer tasks, with a public category score of 57/100. Higher-ranked alternatives are available for workloads where this category decides the choice. Check the underlying rows before treating the aggregate as a workload guarantee.
Does DeepSeek V4 Pro 0813 have full benchmark coverage on BenchLM?
No. DeepSeek V4 Pro 0813 currently has 51 source-displayable rows across 414 tracked benchmark slots. The profile exposes published, non-generated evidence and leaves missing categories blank until an exact evaluation is available. Coverage describes how much was measured; it is not a penalty added to an individual benchmark result.
What is the context window size of DeepSeek V4 Pro 0813?
DeepSeek V4 Pro 0813 has a documented context window of 1M. That figure is the maximum combined prompt and retained-conversation space reported for this exact model; it is not the maximum output length. The profile keeps output limits separate because providers often publish those limits independently.
Last updated August 13, 2026. Runtime fields remain blank until a sourced snapshot exists.
Watch DeepSeek V4 Pro 0813 in the weekly brief
Get one weekly email when material rank, price, availability, or benchmark evidence changes are worth revisiting.
Read a sample issueJoin 2,000+ readers.