ZMIME
Comparison · 3 models · Updated Nov 18, 2026

Granite 4.2 8B vs Perceptron Mk1.5 vs Nemotron 3.5 Lightning 30B A3B

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. IBM

    Granite 4.2 8B

    Released Aug 25, 2026

    65/100
    • ECI—
    • Price$0.06 / $0.15
    • Context131K
  2. Perceptron

    Perceptron Mk1.5

    Released Sep 25, 2026New

    57/100
    • ECI—
    • Price$0.154 / $1.54
    • Context37K
  3. NVIDIA

    Nemotron 3.5 Lightning 30B A3B

    Released Aug 11, 2026

    69/100
    • ECI—
    • Price$0.06 / $0.20
    • Context262K
01 — Verdict

Capability evidence is incomplete

There is not enough shared capability evidence to name an overall winner. The scores (Nemotron 3.5 Lightning 30B A3B 69/100, Granite 4.2 8B 65/100, Perceptron Mk1.5 57/100) 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 priceGranite 4.2 8BGranite 4.2 8B $0.083 · Nemotron 3.5 Lightning 30B A3B $0.095 · Perceptron Mk1.5 $0.502 per 1M tokens (3:1 blend)
  • Longest contextNemotron 3.5 Lightning 30B A3BNemotron 3.5 Lightning 30B A3B 262,144 · Granite 4.2 8B 131,072 · Perceptron Mk1.5 36,864 tokens
  • Widest inputsPerceptron Mk1.5Granite 4.2 8B: Text · Perceptron Mk1.5: Text, Images, Audio, Video · Nemotron 3.5 Lightning 30B A3B: Text
  • Self-hostingGranite 4.2 8B and Nemotron 3.5 Lightning 30B A3BPublishes downloadable weights
How the score is built
MeasureWeightGranite 4.2 8BPerceptron Mk1.5Nemotron 3.5 Lightning 30B A3B
Price50%976194
Inputs & features30%358045
Context window20%321343
Specs & price only100%65/10057/10069/100

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.

Granite 4.2 8B vs Perceptron Mk1.5 vs Nemotron 3.5 Lightning 30B A3B specifications side by side
SpecificationGranite 4.2 8BIBMPerceptron Mk1.5PerceptronNemotron 3.5 Lightning 30B A3BNVIDIA
Capability
Capabilities Index (ECI)———
ECI rank———
Price per million tokens
Input$0.06 (best)$0.154$0.06 (best)
Output$0.15 (best)$1.54$0.20
Cached input———
Blended (3:1)$0.083 (best)$0.502$0.095
Long-context rateSame rateSame rateSame rate
Price sourceMedian of 7 providersMedian of 2 providersMedian of 9 providers
Limits
Context window131,072 tokens36,864 tokens262,144 tokens (best)
Max output117,964 tokens8,192 tokens262,144 tokens (best)
Inputs and features
TextYesYesYes
ImagesNoYesNo
PDFsNoNoNo
AudioNoYesNo
VideoNoYesNo
ReasoningYesYesYes
Tool callingYesYesYes
Structured outputNoNoYes
Availability
WeightsOpenProprietaryOpen
API model ID——nvidia/nemotron-3.5-lightning-30b-a3b
API providers7215 (best)
ReleasedAug 25, 2026Sep 25, 2026Aug 11, 2026
Knowledge cutoff——Sep 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.

  • Granite 4.2 8B$0.90
  • Perceptron Mk1.5$4.63
  • Nemotron 3.5 Lightning 30B A3B$1.00
04 — Questions

Which should you choose?

Which is better: Granite 4.2 8B, Perceptron Mk1.5 or Nemotron 3.5 Lightning 30B A3B?

There is not enough shared capability evidence to name an overall winner. The scores (Nemotron 3.5 Lightning 30B A3B 69/100, Granite 4.2 8B 65/100, Perceptron Mk1.5 57/100) 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, Granite 4.2 8B, Perceptron Mk1.5 or Nemotron 3.5 Lightning 30B A3B?

Granite 4.2 8B is cheaper at $0.06 input / $0.15 output per million tokens (median across 7 API providers). Nemotron 3.5 Lightning 30B A3B costs $0.06 input / $0.20 output per million tokens (median across 9 API providers; free on Nvidia); Perceptron Mk1.5 costs $0.154 input / $1.54 output per million tokens (median across 2 API providers). At a typical mix of three input tokens to one output token, that is $0.083 per million tokens for Granite 4.2 8B versus $0.095 for Nemotron 3.5 Lightning 30B A3B (1.2× as much) and $0.502 for Perceptron Mk1.5 (6.1× as much).

Which scores higher on benchmarks?

There is no independent benchmark that covers all three models yet. Granite 4.2 8B has not been scored yet, Perceptron Mk1.5 has not been scored yet and Nemotron 3.5 Lightning 30B A3B has not been scored yet.

Which is better for coding?

There are no published SWE-bench Verified results for Granite 4.2 8B, Perceptron Mk1.5 and Nemotron 3.5 Lightning 30B A3B 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?

Nemotron 3.5 Lightning 30B A3B has the largest context window at 262,144 tokens, against 131,072 for Granite 4.2 8B and 36,864 for Perceptron Mk1.5. Maximum output per response: Granite 4.2 8B up to 117,964, Perceptron Mk1.5 up to 8,192, Nemotron 3.5 Lightning 30B A3B up to 262,144 tokens.

Which can read images, PDFs, audio or video?

Granite 4.2 8B accepts text; Perceptron Mk1.5 accepts text, images, audio and video; Nemotron 3.5 Lightning 30B A3B accepts text. Perceptron Mk1.5 handles the widest range of inputs.

Are any of these open source?

Granite 4.2 8B and Nemotron 3.5 Lightning 30B A3B publishes its weights and can be self-hosted; Perceptron Mk1.5 is proprietary.

Which is newer?

Perceptron Mk1.5 is the newest, released Sep 25, 2026. Granite 4.2 8B came out Aug 25, 2026; Nemotron 3.5 Lightning 30B A3B came out Aug 11, 2026. Knowledge cutoff: Nemotron 3.5 Lightning 30B A3B Sep 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.