# DeepSeek V3.2

DeepSeek · Open Weight · rank 153 · bench-align-v5

> Every figure below is reproduced as its upstream source published it: nothing is modelled, estimated, interpolated or converted. `Unavailable` means no source published the value — it is never a zero. Each value carries its evidence state and the date it was observed.

Page: https://modelscale.dev/models/deepseek-v3-2  
JSON: https://modelscale.dev/api/model/deepseek-v3-2

## Facts

| Field | Value | Evidence |
| --- | --- | --- |
| Canonical id | `deepseek-v3-2` | — |
| Overall score | 40 | Observed 2026-11-21 · source benchlm:models |
| Verified score (BenchLM): a separate BenchLM figure, not the overall score | 60 | Observed 2026-11-21 · source benchlm:models |
| Context window | 128K tokens | Observed 2026-11-21 · source benchlm:models |
| Release date | 2025-12-01 | Observed 2026-11-21 · source benchlm:models |
| Access type | Open Weight | — |
| Blended $/1M (75% input / 25% output) | $0.32 | Derived from the input and output rates below |

## Capability evidence

Seven axes from the ranking source. An axis the source did not score is unavailable, not zero. Beside each score is BenchLM's separate verified score for that axis; the two are published side by side and are not interchangeable.

| Axis | Score | Evidence | Verified (BenchLM) | Verified evidence |
| --- | --- | --- | --- | --- |
| Agentic | Unavailable | Unavailable · source benchlm:models | Unavailable | Unavailable · source benchlm:models |
| Coding | 69.6 | Observed 2026-11-21 · source benchlm:models | 69.5 | Observed 2026-11-21 · source benchlm:models |
| Knowledge | 41.1 | Observed 2026-11-21 · source benchlm:models | 41.1 | Observed 2026-11-21 · source benchlm:models |
| Reasoning | 55.5 | Observed 2026-11-21 · source benchlm:models | 55.5 | Observed 2026-11-21 · source benchlm:models |
| Multimodal & Grounded | Unavailable | Unavailable · source benchlm:models | Unavailable | Unavailable · source benchlm:models |
| Instruction Following | 58.1 | Observed 2026-11-21 · source benchlm:models | 58.1 | Observed 2026-11-21 · source benchlm:models |
| Math | 40 | Last good 2026-10-08 · source benchlm:models | 40 | Last good 2026-10-08 · source benchlm:models |

## Runtime service evidence

Measured values with the date they were observed. Nothing is inferred from a sibling model or a provider claim. Regional or per-endpoint measurements appear only when the API supplies them; none are modelled.

| Measurement | Value | Observed | Last good | Evidence |
| --- | --- | --- | --- | --- |
| Time to first token | 3.75 s | 2026-11-21 | 2026-11-21 | Observed 2026-11-21 · source benchlm:speed |
| Throughput | 35 tok/s | 2026-11-21 | 2026-11-21 | Observed 2026-11-21 · source benchlm:speed |

## Endpoint and price matrix

Every published price component, including cache reads and writes.

| Component | USD | Evidence |
| --- | --- | --- |
| Input / 1M tokens | $0.28 | Observed 2026-11-21 · source benchlm:pricing |
| Output / 1M tokens | $0.42 | Observed 2026-11-21 · source benchlm:pricing |
| Cache read / 1M tokens | $0.028 | Observed 2026-11-21 · source benchlm:pricing |
| Cache write / 1M tokens | Unavailable | Unavailable · source openrouter:pricing |
| Blended / 1M (75% input / 25% output) | $0.32 | Derived — from the input and output rates above; it has no source record of its own |
| Cost per successful task (LiveBench) | Unavailable | Unavailable · source livebench:table |

**Pricing note.** Historical DeepSeek V3.2 pricing for the former `deepseek-chat` alias: $0.28 input (cache miss) / $0.028 cache-hit input / $0.42 output per million tokens. DeepSeek's published July 24, 2026 deprecation date has passed; current requests should use the V4 model IDs.

## Workload-aware monthly cost example

10 conversations per day × 8 messages × 22 active days, 1200 input and 400 output tokens per message, no cache. Derived here from the published rates above by this site's own calculator — not a figure any source published.

| Field | Value |
| --- | --- |
| Modelled monthly cost | $0.89 |
| Modelled tokens | 2.82M |

## Benchmark record

21 matched benchmark rows with their published value, unit, and provenance.

### Knowledge

| Benchmark | Value | Tasks | Difficulty | Provenance |
| --- | --- | --- | --- | --- |
| Artificial Analysis Intelligence Index | 16.0 | Cross-benchmark intelligence index | Display-only external reference | [Artificial Analysis](https://artificialanalysis.ai/) |
| AA-GPQA Diamond (Artificial Analysis GPQA Diamond) | 75.1 | Graduate-level science questions | Graduate-level science reasoning | [Artificial Analysis GPQA Diamond Benchmark Leaderboard](https://artificialanalysis.ai/evaluations/gpqa-diamond) |
| AA-HLE (Artificial Analysis Humanity's Last Exam) | 11.2 | Expert-level questions | Frontier expert reasoning | [Artificial Analysis Humanity's Last Exam Benchmark Leaderboard](https://artificialanalysis.ai/evaluations/hle) |
| AA-Omniscience Index (Artificial Analysis Omniscience Index) | -46.9 | Knowledge questions | Broad factual knowledge | [AA-Omniscience: Knowledge and Hallucination Benchmark](https://artificialanalysis.ai/evaluations/omniscience) |
| AA-Omniscience Accuracy (Artificial Analysis Omniscience Accuracy) | 24.0 | Knowledge questions | Broad knowledge | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/grok-4-3) |
| AA-Omniscience Hallucination Rate (Artificial Analysis Omniscience Hallucination Rate) | 93.3 | Knowledge questions | Factuality | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/grok-4-3) |

### Coding

| Benchmark | Value | Tasks | Difficulty | Provenance |
| --- | --- | --- | --- | --- |
| SWE-Rebench | 60.9 | Fresh GitHub issues (rolling window) | Professional software engineering | [SWE-Rebench: Contamination-Free Evaluation of Software Engineering Agents](https://swe-rebench.com/) |
| React Native Evals | 71.5 | React Native app implementation tasks | Production mobile app engineering | [React Native Evals](https://rn-evals.vercel.app/) |

### Mathematics

| Benchmark | Value | Tasks | Difficulty | Provenance |
| --- | --- | --- | --- | --- |
| FrontierMath v2 (Tiers 1-3) (FrontierMath v2 Tiers 1-3) | 22.100 | 285 private advanced mathematics problems | From olympiad-plus to early research mathematics | [FrontierMath v2 benchmark hub](https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2) |
| FrontierMath v2 (Tier 4) (FrontierMath v2 Tier 4) | 2.100 | 41 private extreme-difficulty mathematics problems | Research-level mathematics requiring hours or days of expert work | [FrontierMath Tier 4 v2 leaderboard](https://epoch.ai/benchmarks/frontiermath-tier-4-v2?view=graph&tab=leaderboard) |

### Reasoning

| Benchmark | Value | Tasks | Difficulty | Provenance |
| --- | --- | --- | --- | --- |
| AA-LCR (Artificial Analysis Long Context Reasoning) | 45.7 | Long-context reasoning tasks | Long-context reasoning | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/grok-4-3) |
| CritPt (Critical Physics Tasks) | 0.9 | Research-level physics questions | Research-level physics reasoning | [CritPt Benchmark Leaderboard](https://artificialanalysis.ai/evaluations/critpt) |

### Instruction Following

| Benchmark | Value | Tasks | Difficulty | Provenance |
| --- | --- | --- | --- | --- |
| AA-IFBench (Artificial Analysis IFBench) | 49.0 | Verifiable instruction constraints | Instruction precision | [Artificial Analysis IFBench Benchmark Leaderboard](https://artificialanalysis.ai/evaluations/ifbench) |

### Agentic

| Benchmark | Value | Tasks | Difficulty | Provenance |
| --- | --- | --- | --- | --- |
| Gert Labs (Gert Labs Composite Game Benchmark) | 29.57 | Novel game environments | Agentic coding and decision-making | [Gert Labs rankings](https://gertlabs.com/rankings) |
| τ²-bench results (τ²-Bench Tool-Agent-User Evaluation) | 78.9 | Airline, retail, and telecom customer-service task sets | Dual-control customer-service workflows | [τ²-Bench: Evaluating Conversational Agents in a Dual-Control Environment](https://arxiv.org/abs/2506.07982) |
| Claw-Eval | 40.2 | 300 tasks, 2,159 rubrics | Real-world general, multi-turn, and native multimodal agent execution | [Claw-Eval: Towards Trustworthy Evaluation of Autonomous Agents](https://arxiv.org/abs/2604.06132) |
| VITA-Bench | 18.5 | Interactive consumer-service agent tasks | Long-horizon real-world workflows | [VitaBench: Benchmarking LLM Agents with Versatile Interactive Tasks in Real-world Applications](https://vitabench.github.io/) |

### Multimodal & Grounded

| Benchmark | Value | Tasks | Difficulty | Provenance |
| --- | --- | --- | --- | --- |
| Design Arena Website (Design Arena Website Elo) | 1164 | Website generation comparisons | Design and website generation | [OpenRouter Grok 4.3 benchmarks](https://openrouter.ai/x-ai/grok-4.3/benchmarks) |

### external

| Benchmark | Value | Tasks | Difficulty | Provenance |
| --- | --- | --- | --- | --- |
| BinaryAudit (BinaryAudit (backdoor detection in compiled binaries)) | 12.3 | 19 backdoor and timebomb detection tasks on stripped binaries, three runs each (57 runs); 6 tooling and 8 clean-binary tasks reported alongside | Reverse engineering without source code or symbols | [We hid backdoors in ~40 MB binaries and asked AI + Ghidra to find them (Quesma)](https://quesma.com/blog/introducing-binaryaudit/) |
| OTelBench (OTelBench (OpenTelemetry instrumentation tasks)) | 11.6 | 23 OpenTelemetry instrumentation tasks in 11 languages, three attempts each (69 runs) | Distributed tracing and context propagation across services | [Benchmarking OpenTelemetry: can AI trace your failed login? (Quesma)](https://quesma.com/blog/introducing-otel-bench/) |
| NYT Connections (NYT Connections LLM Benchmark (extended version)) | 8.2 | 940 NYT Connections puzzles, each with up to four added trick words | Wordplay, word knowledge and decoys that fit almost-but-not-quite | [NYT Connections LLM Benchmark (Lech Mazur, GitHub)](https://github.com/lechmazur/nyt-connections) |

## Lifecycle and limitations log

Lifecycle events the source associates with this model.

No lifecycle event references this model. That is not evidence the model has no lifecycle plan — only that this source published none.

## What this profile does not claim

Values are reproduced exactly as their sources published them, in the units those sources declared; none are converted, interpolated, or averaged across providers. Any field marked unavailable was attempted and not returned. Last attempted fetch for this model's score: 2026-11-21 01:17 UTC.
