Skip to main content

    DeepSeek

    DeepSeek V3.2

    DeepSeek V3.2 is a historical text model using DeepSeek Sparse Attention; its published weights remain available. By the April 24, 2026 V4 announcement, the deepseek-chat and deepseek-reasoner aliases already routed to V4 Flash, and those aliases were scheduled to end after July 24 at 15:59 UTC. Separately, Alibaba Cloud Model Studio schedules its hosted deepseek-v3.2 endpoint for retirement on October 10, 2026 at 00:00 UTC+8, subject to rollout timing.

    128K tokens
    Text only
    Pro
    Provider
    DeepSeek
    Context
    128K
    Catalog · As of Aug 11, 2026Stale · 31d old
    yno subscription tier
    Pro
    Released
    Dec 1, 2025
    Speed
    Fast
    Reasoning
    Advanced
    Modality
    Text only

    Benchmarks

    Quality

    AA Intelligence Index
    16points
    Artificial Analysis · Date unknownAge unknown
    LiveCodeBench
    59.3%
    Artificial Analysis · Date unknownAge unknown
    GPQA Diamond
    75.1%
    Artificial Analysis · Date unknownAge unknown

    Speed and latency

    Output speed
    102tokens/s
    Artificial Analysis · As of Aug 11, 2026Stale · 31d old

    API pricing

    API input cost
    $0.28/ 1M tokens
    Artificial Analysis · Date unknownAge unknown
    API output cost
    $0.42/ 1M tokens
    Artificial Analysis · Date unknownAge unknown

    Benchmark variant: DeepSeek V3.2 (Non-reasoning)

    AA data retrieved Sep 11, 2026 · Artificial Analysis

    Indexes use points; evaluations use accuracy percentages. API measurements do not measure yno application performance. Methodology

    Capabilities

    • Long Context
    • Cost-Effective
    • Document Processing
    • Analysis

    Use cases

    • Document analysis
    • Long-form content
    • Research
    • Budget projects

    Strengths

    • Published weights
    • DeepSeek Sparse Attention
    • Historical text-model comparison
    • Third-party hosting has separate lifecycles

    Best for

    Historical comparison or deployments using the published V3.2 weights

    Use DeepSeek V3.2 in yno.ai

    No credit card required

    Related models

    Frequently asked

    What is DeepSeek V3.2?
    DeepSeek V3.2 is a historical text model using DeepSeek Sparse Attention; its published weights remain available. By the April 24, 2026 V4 announcement, the deepseek-chat and deepseek-reasoner aliases already routed to V4 Flash, and those aliases were scheduled to end after July 24 at 15:59 UTC. Separately, Alibaba Cloud Model Studio schedules its hosted deepseek-v3.2 endpoint for retirement on October 10, 2026 at 00:00 UTC+8, subject to rollout timing.
    How much does DeepSeek V3.2 cost in yno.ai?
    DeepSeek V3.2 is available on the Pro tier of yno.ai.
    What can DeepSeek V3.2 do?
    DeepSeek V3.2 is best for Historical comparison or deployments using the published V3.2 weights. Its main capabilities include Long Context, Cost-Effective, Document Processing, Analysis.
    How does DeepSeek V3.2 compare to other models?
    DeepSeek V3.2 excels at Published weights and is recommended for Document analysis, Long-form content, Research, Budget projects. See related models below.
    How do I use DeepSeek V3.2 in yno.ai?
    Sign up for yno.ai, select DeepSeek V3.2 from the model picker, and start chatting.