# DeepSeek-V4-Pro

DeepSeek-V4-Pro is an AI model released by DeepSeek on Apr 24 2026. It has 1.6T parameters, a 1M context window and open weight. Tracked results include 93.5% on LiveCodeBench, 14% on BullshitBench v2 and 90.1% on GPQA Diamond.

## Facts

| Field | Value |
| --- | --- |
| Model | DeepSeek-V4-Pro |
| Developer | DeepSeek |
| Release date | Friday, Apr 24 2026 |
| Licensing | Open Weight |
| Parameters | 1.6T |
| Context window | 1M |

## Tracked benchmark scores

| Benchmark | Score | Source | What it measures |
| --- | --- | --- | --- |
| CheatBench | 75.1% | [CheatBench](https://cheatbench.ai/), retrieved 2026-09-25 | Measures how often AI agents try to cheat on difficult assignments, such as reading hidden answers, copying work or manipulating grading. The overall score gives equal weight to ten categories; the sycophancy category measures how far an agent shifts its beliefs toward a user's stated views. Scores describe each model in its tested agent setup, and task success is measured separately. Lower is better. |
| BullshitBench v2 | 14% | [BullshitBench](https://github.com/petergpt/bullshit-benchmark) | Given a confidently-worded but nonsensical prompt, does the AI spot that it makes no sense and push back — instead of playing along and inventing an answer? The score is how often it clearly called out the nonsense. Higher is better. |
| LiveCodeBench | 93.5% | [BenchLM](https://benchlm.ai), retrieved 2026-07-11 | Coding problems published so recently the AI can't have seen them in training — a contamination-free test of raw programming skill. Higher is better. |
| BrowseComp | 83.4% | [BenchLM](https://benchlm.ai), retrieved 2026-07-11 | Can the AI browse the web and track down hard-to-find answers? Higher is better. |
| GPQA Diamond | 90.1% | Lab | Graduate-level science questions in biology, physics, and chemistry — hard enough that subject-matter PhDs score around 65%. Higher is better. |

## About DeepSeek-V4-Pro

DeepSeek V4 Pro was introduced in the April 2026 V4 preview alongside V4 Flash. Pro used a mixture-of-experts architecture with 1.6 trillion total parameters and about 49 billion active per token. The preview supported a one-million-token context and offered thinking and non-thinking modes.

The V4 generation combined token-wise compression with DeepSeek Sparse Attention to reduce the cost of processing long inputs. Pro was available through DeepSeek’s hosted services and as downloadable weights. This entry represents the original preview checkpoint; the later V4-Pro-0813 release was a separate update, so its benchmarks and reasoning controls should not be assigned to the April build.

## Questions and answers

### When was DeepSeek-V4-Pro released?

DeepSeek-V4-Pro was released by DeepSeek on Friday, Apr 24 2026.

### Who made DeepSeek-V4-Pro?

DeepSeek-V4-Pro was built by DeepSeek. Chinese AI lab known for efficient, open-weight models. Gained attention for strong performance at lower cost.

### What benchmark scores did DeepSeek-V4-Pro get?

DeepSeek-V4-Pro reports 5 tracked benchmark scores — CheatBench: 75.1%; BullshitBench v2: 14%; LiveCodeBench: 93.5%; BrowseComp: 83.4%; GPQA Diamond: 90.1%. Tracked scores may come from lab reports or independent benchmarks; source details accompany the benchmark data. It holds the best score among all models tracked here on LiveCodeBench.

### What is the context window of DeepSeek-V4-Pro?

DeepSeek-V4-Pro has a context window of 1M. Providers may specify separate input and output token limits.

### How many parameters does DeepSeek-V4-Pro have?

DeepSeek-V4-Pro is reported at 1.6T parameters.

### Is DeepSeek-V4-Pro open source?

Partly. DeepSeek-V4-Pro is an open-weight model: the trained weights are free to download and run locally or on your own infrastructure, but the training data and code are not fully released and the license may restrict some commercial uses. It is not open source in the strict sense.

### What came before and after DeepSeek-V4-Pro?

DeepSeek's previous tracked release was DeepSeek-V3.2 on Dec 1 2025, 144 days earlier. It was followed by DeepSeek-V4-Flash on Apr 24 2026.


---

Canonical page: https://aireleasetracker.com/model/deepseek/deepseek-v4-pro
Site index: https://aireleasetracker.com/llms.txt
Source: AI Release Tracker (https://aireleasetracker.com). Most benchmark scores come from lab launch material; gathered results identify the leaderboard that published them.