ZMIME
Comparison · 3 models · Updated Nov 18, 2026

Qwen3.8 2.4T A95B vs DeepSeek V3.2 vs Fugu Max

No independent capability benchmark covers every model here yet. The score uses price, inputs & features and context window.

Price in rating: Included · Exclude price

  1. Alibaba (Qwen)

    Qwen3.8 2.4T A95B

    Released Aug 12, 2026

    32/100
    • ECI—
    • Price$2.00 / $6.00
    • Context262K
  2. DeepSeek

    DeepSeek V3.2

    Released Dec 1, 2025

    52/100
    • ECI146.3
    • Price$0.28 / $0.42
    • Context128K
  3. Sakana AI

    Fugu Max

    Released Sep 11, 2026

    47/100
    • ECI—
    • Price$2.00 / $6.00
    • Context1M
01 — Verdict

Capability evidence is incomplete

There is not enough shared capability evidence to name an overall winner. The scores (DeepSeek V3.2 52/100, Fugu Max 47/100, Qwen3.8 2.4T A95B 32/100) compare the available specifications and price. The score weighs price 50%, inputs & features 30%, context window 20%. None of the shared benchmarks cover every model here yet, so capability is left out of this verdict.

  • CapabilityNot enough dataNo independent benchmark covers every model here yet
  • Lowest priceDeepSeek V3.2DeepSeek V3.2 $0.315 · Qwen3.8 2.4T A95B $3.00 · Fugu Max $3.00 per 1M tokens (3:1 blend)
  • Longest contextFugu MaxFugu Max 1,000,000 · Qwen3.8 2.4T A95B 262,144 · DeepSeek V3.2 128,000 tokens
  • Widest inputsFugu MaxQwen3.8 2.4T A95B: Text · DeepSeek V3.2: Text · Fugu Max: Text, Images, PDFs
  • Self-hostingQwen3.8 2.4T A95B and DeepSeek V3.2Publishes downloadable weights (qwen3.8-max and MIT License)
How the score is built
MeasureWeightQwen3.8 2.4T A95BDeepSeek V3.2Fugu Max
Price50%267126
Inputs & features30%353570
Context window20%433264
Specs & price only100%32/10052/10047/100

Left out because at least one model lacks the data: capability. The remaining weights were rescaled.

02 — Side by side

Every spec in one table

Highlighted cells lead their row. Dashes mean the data is not published.

Qwen3.8 2.4T A95B vs DeepSeek V3.2 vs Fugu Max specifications side by side
SpecificationQwen3.8 2.4T A95BAlibaba (Qwen)DeepSeek V3.2DeepSeekFugu MaxSakana AI
Capability
Capabilities Index (ECI)—146.3—
ECI rank—#87 of 158—
GPQA DiamondGraduate-level science questions—83.4%—
OTIS Mock AIME 2024–2025Competition mathematics—87.8%—
Price per million tokens
Input$2.00$0.28 (best)$2.00
Output$6.00$0.42 (best)$6.00
Cached input———
Blended (3:1)$3.00$0.315 (best)$3.00
Long-context rateSame rateSame rateSame rate
Price sourceMedian of 26 providersMedian of 30 providersMedian of 9 providers
Limits
Context window262,144 tokens128,000 tokens1,000,000 tokens (best)
Max output131,072 tokens (best)64,000 tokens128,000 tokens
Inputs and features
TextYesYesYes
ImagesNoNoYes
PDFsNoNoYes
AudioNoNoNo
VideoNoNoNo
ReasoningYesYesYes
Tool callingYesYesYes
Structured outputNoNoNo
Availability
WeightsOpenqwen3.8-maxOpenMIT LicenseProprietary
API model ID———
API providers2630 (best)9
ReleasedAug 12, 2026Dec 1, 2025Sep 11, 2026
Knowledge cutoff—Jul 2024—
03 — Cost

What would a month cost?

Enter your expected volume in millions of tokens. List prices only; caching and batch discounts would lower these.

  • Qwen3.8 2.4T A95B$32.00
  • DeepSeek V3.2$3.64
  • Fugu Max$32.00
04 — Questions

Which should you choose?

Which is better: Qwen3.8 2.4T A95B, DeepSeek V3.2 or Fugu Max?

There is not enough shared capability evidence to name an overall winner. The scores (DeepSeek V3.2 52/100, Fugu Max 47/100, Qwen3.8 2.4T A95B 32/100) compare the available specifications and price. The score weighs price 50%, inputs & features 30%, context window 20%. None of the shared benchmarks cover every model here yet, so capability is left out of this verdict.

Which is cheaper, Qwen3.8 2.4T A95B, DeepSeek V3.2 or Fugu Max?

DeepSeek V3.2 is cheaper at $0.28 input / $0.42 output per million tokens (median across 30 API providers). Qwen3.8 2.4T A95B costs $2.00 input / $6.00 output per million tokens (median across 26 API providers); Fugu Max costs $2.00 input / $6.00 output per million tokens (median across 9 API providers). At a typical mix of three input tokens to one output token, that is $0.315 per million tokens for DeepSeek V3.2 versus $3.00 for Qwen3.8 2.4T A95B (9.5× as much) and $3.00 for Fugu Max (9.5× as much).

Which scores higher on benchmarks?

There is no independent benchmark that covers all three models yet. Qwen3.8 2.4T A95B has not been scored yet, DeepSeek V3.2 has an ECI of 146.3 and Fugu Max has not been scored yet.

Which is better for coding?

There are no published SWE-bench Verified results for Qwen3.8 2.4T A95B, DeepSeek V3.2 and Fugu Max yet, so there is no like-for-like coding score. Test both on a sample of your own tasks. All three support tool calling for agent workflows.

Which has the bigger context window?

Fugu Max has the largest context window at 1,000,000 tokens, against 262,144 for Qwen3.8 2.4T A95B and 128,000 for DeepSeek V3.2. Maximum output per response: Qwen3.8 2.4T A95B up to 131,072, DeepSeek V3.2 up to 64,000, Fugu Max up to 128,000 tokens.

Which can read images, PDFs, audio or video?

Qwen3.8 2.4T A95B accepts text; DeepSeek V3.2 accepts text; Fugu Max accepts text, images and PDFs. Fugu Max handles the widest range of inputs.

Are any of these open source?

Qwen3.8 2.4T A95B and DeepSeek V3.2 publishes its weights (qwen3.8-max and MIT License) and can be self-hosted; Fugu Max is proprietary.

Which is newer?

Fugu Max is the newest, released Sep 11, 2026. Qwen3.8 2.4T A95B came out Aug 12, 2026; DeepSeek V3.2 came out Dec 1, 2025. Knowledge cutoff: DeepSeek V3.2 Jul 2024.

How do you decide the winner?

This rating uses price (50%, blended price per million tokens (3 input : 1 output), log scale); inputs & features (30%, image, PDF, audio and video input, tool calling, structured output and reasoning); context window (20%, maximum tokens per request, log scale). Dimensions missing for any model are dropped and the remaining weights rescaled, so every model is judged on the same evidence. Specs and prices come from public model listings and the labs’ own API pages; capability scores come from independent benchmark runs. Data updated Nov 18, 2026.