ZMIME
Comparison · 3 models · Updated Nov 18, 2026

Step 3.5 Flash 2603 vs Perceptron Mk1 vs Gemma 4 E4B IT

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. StepFun

    Step 3.5 Flash 2603

    Released Apr 2, 2026

    65/100
    • ECI—
    • Price$0.10 / $0.30
    • Context256K
  2. Perceptron

    Perceptron Mk1

    Released May 12, 2026

    47/100
    • ECI—
    • Price$0.15 / $1.50
    • Context33K
  3. Google

    Gemma 4 E4B IT

    Released Apr 2, 2026

    70/100
    • ECI—
    • Price$0.085 / $0.35
    • Context131K
01 — Verdict

Capability evidence is incomplete

There is not enough shared capability evidence to name an overall winner. The scores (Gemma 4 E4B IT 70/100, Step 3.5 Flash 2603 65/100, Perceptron Mk1 47/100) 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 priceStep 3.5 Flash 2603Step 3.5 Flash 2603 $0.15 · Gemma 4 E4B IT $0.151 · Perceptron Mk1 $0.487 per 1M tokens (3:1 blend)
  • Longest contextStep 3.5 Flash 2603Step 3.5 Flash 2603 256,000 · Gemma 4 E4B IT 131,072 · Perceptron Mk1 32,768 tokens
  • Widest inputsPerceptron Mk1 and Gemma 4 E4B ITStep 3.5 Flash 2603: Text · Perceptron Mk1: Text, Images, Video · Gemma 4 E4B IT: Text, Images, Audio
  • Self-hostingStep 3.5 Flash 2603 and Gemma 4 E4B ITPublishes downloadable weights
How the score is built
MeasureWeightStep 3.5 Flash 2603Perceptron Mk1Gemma 4 E4B IT
Price50%856285
Inputs & features30%454570
Context window20%431132
Specs & price only100%65/10047/10070/100

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.

Step 3.5 Flash 2603 vs Perceptron Mk1 vs Gemma 4 E4B IT specifications side by side
SpecificationStep 3.5 Flash 2603StepFunPerceptron Mk1PerceptronGemma 4 E4B ITGoogle
Capability
Capabilities Index (ECI)———
ECI rank———
Price per million tokens
Input$0.10$0.15$0.085 (best)
Output$0.30 (best)$1.50$0.35
Cached input$0.02——
Blended (3:1)$0.15 (best)$0.487$0.151
Long-context rateSame rateSame rateSame rate
Price sourceOfficial StepFun (Global) APIMedian of 3 providersMedian of 2 providers
Limits
Context window256,000 tokens (best)32,768 tokens131,072 tokens
Max output256,000 tokens (best)8,192 tokens8,192 tokens
Inputs and features
TextYesYesYes
ImagesNoYesYes
PDFsNoNoNo
AudioNoNoYes
VideoNoYesNo
ReasoningYeslow · highYesYes
Tool callingYesNoYes
Structured outputYesNoNo
Availability
WeightsOpenProprietaryOpen
API model IDstep-3.5-flash-2603——
API providers333
ReleasedApr 2, 2026May 12, 2026Apr 2, 2026
Knowledge cutoffJan 2025——
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.

  • Step 3.5 Flash 2603$1.60
  • Perceptron Mk1$4.50
  • Gemma 4 E4B IT$1.55
04 — Questions

Which should you choose?

Which is better: Step 3.5 Flash 2603, Perceptron Mk1 or Gemma 4 E4B IT?

There is not enough shared capability evidence to name an overall winner. The scores (Gemma 4 E4B IT 70/100, Step 3.5 Flash 2603 65/100, Perceptron Mk1 47/100) 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, Step 3.5 Flash 2603, Perceptron Mk1 or Gemma 4 E4B IT?

Step 3.5 Flash 2603 is cheaper at $0.10 input / $0.30 output per million tokens (official StepFun (Global) API price). Gemma 4 E4B IT costs $0.085 input / $0.35 output per million tokens (median across 2 API providers); Perceptron Mk1 costs $0.15 input / $1.50 output per million tokens (median across 3 API providers). At a typical mix of three input tokens to one output token, that is $0.15 per million tokens for Step 3.5 Flash 2603 versus $0.151 for Gemma 4 E4B IT (1× as much) and $0.487 for Perceptron Mk1 (3.3× as much).

Which scores higher on benchmarks?

There is no independent benchmark that covers all three models yet. Step 3.5 Flash 2603 has not been scored yet, Perceptron Mk1 has not been scored yet and Gemma 4 E4B IT has not been scored yet.

Which is better for coding?

There are no published SWE-bench Verified results for Step 3.5 Flash 2603, Perceptron Mk1 and Gemma 4 E4B IT yet, so there is no like-for-like coding score. Test both on a sample of your own tasks. Note that Perceptron Mk1 does not support tool calling, which most coding agents need.

Which has the bigger context window?

Step 3.5 Flash 2603 has the largest context window at 256,000 tokens, against 131,072 for Gemma 4 E4B IT and 32,768 for Perceptron Mk1. Maximum output per response: Step 3.5 Flash 2603 up to 256,000, Perceptron Mk1 up to 8,192, Gemma 4 E4B IT up to 8,192 tokens.

Which can read images, PDFs, audio or video?

Step 3.5 Flash 2603 accepts text; Perceptron Mk1 accepts text, images and video; Gemma 4 E4B IT accepts text, images and audio. Perceptron Mk1 handles the widest range of inputs.

Are any of these open source?

Step 3.5 Flash 2603 and Gemma 4 E4B IT publishes its weights and can be self-hosted; Perceptron Mk1 is proprietary.

Which is newer?

Perceptron Mk1 is the newest, released May 12, 2026. Step 3.5 Flash 2603 came out Apr 2, 2026; Gemma 4 E4B IT came out Apr 2, 2026. Knowledge cutoff: Step 3.5 Flash 2603 Jan 2025.

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.