gpt.buzz
Sign in

Models/DeepSeek

DeepSeek logoDeepSeek-V4-Pro

DeepSeek · DeepSeekreleased 2026-04-22open sourceupdated 24 days ago

DeepSeek's flagship open-weight MoE. 1.6T parameters with 49B activated, 1M-token context, and a hybrid attention scheme (CSA + HCA) that delivers long-context inference at ~27% of V3.2's FLOPs.

Specifications

Context window
1,000,000 tokens
Parameters
1.6T (49B active)
Modality
text
License
MIT
Family
DeepSeek
Release date
2026-04-22

Links

Provider status

DeepSeek APIAll Systems Operational
Last incident: 【已恢复】DeepSeek 网页/API不可用([Resolved]DeepSeek Web/API Service Not Available)about 2 months ago
30-day uptime100.00%
View DeepSeek's status page →

Timeline

  1. Released

    Initial public availability.

  2. Pricing changes, lineage updates, and new benchmark results appear here as they happen. See the releases feed for the latest vendor activity.

API pricing

No API pricing recorded yet for DeepSeek-V4-Pro.

Looking for consumer subscriptions? See DeepSeek's plans →

Benchmarks

Model Index →
BenchmarkScoreSettingMeasuredSource
MMLU-Pro

general knowledge

84.2%5-shot CoT2026-04-22source ↗
GPQA Diamond

reasoning

82.4%0-shot CoT2026-04-22source ↗
HumanEval

coding

95.1%pass@12026-04-22source ↗
Aider Polyglot

coding

80.1%edit + test2026-04-25source ↗
AIME 2025

math

88.6%CoT extended2026-04-22source ↗
LiveCodeBench

coding

72.4%pass@1, post-cutoff2026-04-30source ↗

Want to see how DeepSeek-V4-Pro ranks across these? Open the Model Index leaderboard →

Infrastructure context

All intelligence →

Compute, silicon, and capex events that shape DeepSeek-V4-Pro's economics.

  • Compute clusterdirect

    DeepSeek operates ~50k H800 GPU fleet despite export controls

    DeepSeek's training fleet — believed to be ~50,000 H800 GPUs at peak — was assembled prior to the October 2023 US export controls extension. DeepSeek-V4-Pro's training run reportedly used <8M GPU-hours total, less than 10% of GPT-5's estimated budget, leveraging the hybrid CSA+HCA attention scheme to compress FLOPs.

    Accelerators: 50k · H800
    Location: Hangzhou, China
    SemiAnalysis estimate

Compare DeepSeek-V4-Pro with…

Related news

No tagged articles yet. The aggregator surfaces mentions every 15 minutes.