Monday, October 26, 2026 · Edition updated 6:00 AM ETMON, OCT 26 editionNext edition in Next in --:--:--
Intelligence Verdict
Compare AI models

DeepSeek-V4-Pro-0813 vs Qwen3.8-Max

DeepSeek's DeepSeek-V4-Pro-0813 and Alibaba (Qwen)'s Qwen3.8-Max, side by side: what each costs, how much it can read at once, whether you can run it yourself, and where independent tests place it.

  • DeepSeek-V4-Pro-0813cheaper per output token
  • 1Mcontext window, both
  • 0–1higher placings on boards ranking both
In short

How they differ.

Worked out from the figures below. Prices and specs are the makers' own, read Oct 20, 2026.

  • DeepSeek-V4-Pro-0813 is cheaper per output token: $3.96 against $6 per million tokens, so Qwen3.8-Max costs 1.5 times as much.
  • DeepSeek-V4-Pro-0813 can be downloaded and run on your own machine; Qwen3.8-Max can't.
  • On the 1 independent board that rank both, Qwen3.8-Max places higher on 1 and DeepSeek-V4-Pro-0813 on 0.
Side by side

The specs.

DeepSeek-V4-Pro-0813Qwen3.8-Max
Made byDeepSeek · ChinaAlibaba (Qwen) · China
ReleasedAug 13, 2026Sep 2, 2026
Price per 1M tokens$1.32 in / $3.96 out (peak-hours price)$2 in / $6 out
Context window1M1M
Takes intexttext, image, video
Download and run it yourselfYes (MIT)No
Size1.6T parameters, 49B active2.4T (MoE) parameters, 95B active
Maker says—“A 2.4-trillion-parameter MoE flagship with a major leap in coding and office productivity”
SourcesPrices, Oct 20, 2026Prices, Oct 11, 2026 · Description
Independent tests

Where they place.

Position on each board we follow (1 is best). Epoch AI's and Vectara's scores are shown; the other boards are linked for their own figures.

BoardDeepSeek-V4-Pro-0813Qwen3.8-Max
Epoch Capabilities IndexEpoch AI, Oct 20, 2026Not listed#20156.88
LMArena creative writingLMArena, Oct 15, 2026Not listed#19
LMArena WebDevLMArena, Oct 19, 2026Not listed#11
LMArena mathLMArena, Oct 15, 2026Not listed#18
LMArena visionLMArena, Oct 19, 2026Not listed#2
Humanity's Last ExamScale AI#16Not listed
Vals IndexVals AI, Oct 19, 2026#17Not listed
Terminal-Bench 4.0Vals AI, Oct 19, 2026Not listed#16
GPQA Diamond (run by Vals AI, archived)Vals AI, Sep 1, 2026#17#6
MMMU Pro (run by Vals AI, archived)Vals AI, Sep 1, 2026Not listed#12
Vectara hallucination testVectara, Oct 16, 2026#448.6%Not listed