ZMIME
Comparison · 3 models · Updated Oct 4, 2026

Laguna S 2.1 vs Nemotron 3.5 Lightning 30B A3B vs Gemma 4 12B IT

Gemma 4 12B IT comes out ahead, 75 to 71 and 69 on our weighted score, though Nemotron 3.5 Lightning 30B A3B is 30% cheaper per token.

  1. Poolside

    Laguna S 2.1

    Released Jul 21, 2026

    69/100
    • ECI—
    • Price$0.10 / $0.20
    • Context1.05M
    Full specs Remove Laguna S 2.1

  2. NVIDIA

    Nemotron 3.5 Lightning 30B A3B

    Released Aug 11, 2026

    71/100
    • ECI—
    • Price$0.05 / $0.20
    • Context262K
    Full specs Remove Nemotron 3.5 Lightning 30B A3B
  3. Our pick

    Google

    Gemma 4 12B IT

    Released Jun 9, 2026

    75/100
    • ECI—
    • Price$0.075 / $0.275
    • Context262K
    Full specs Remove Gemma 4 12B IT
01 — Verdict

Gemma 4 12B IT is our pick

Gemma 4 12B IT is the better all-round choice, scoring 75/100 against Nemotron 3.5 Lightning 30B A3B (71) and Laguna S 2.1 (69). It leads on inputs & features. Laguna S 2.1 wins on context window. Nemotron 3.5 Lightning 30B A3B wins on 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 priceNemotron 3.5 Lightning 30B A3BNemotron 3.5 Lightning 30B A3B $0.087 · Laguna S 2.1 $0.125 · Gemma 4 12B IT $0.125 per 1M tokens (3:1 blend)
  • Longest contextLaguna S 2.1Laguna S 2.1 1,048,576 · Nemotron 3.5 Lightning 30B A3B 262,144 · Gemma 4 12B IT 262,144 tokens
  • Widest inputsGemma 4 12B ITLaguna S 2.1: Text · Nemotron 3.5 Lightning 30B A3B: Text · Gemma 4 12B IT: Text, Images
  • Self-hostingAll open weightsEvery model here can be downloaded and run on your own hardware
How the score is built
MeasureWeightLaguna S 2.1Nemotron 3.5 Lightning 30B A3BGemma 4 12B IT
Price50%9310093
Inputs & features30%354570
Context window20%613737
Overall100%69/10071/10075/100

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.

Laguna S 2.1 vs Nemotron 3.5 Lightning 30B A3B vs Gemma 4 12B IT specifications side by side
SpecificationLaguna S 2.1PoolsideNemotron 3.5 Lightning 30B A3BNVIDIAGemma 4 12B ITGoogle
Capability
Capabilities Index (ECI)———
ECI rank———
Price per million tokens
Input$0.10$0.05 (best)$0.075
Output$0.20 (best)$0.20 (best)$0.275
Cached input———
Blended (3:1)$0.125$0.087 (best)$0.125
Long-context rateSame rateSame rateSame rate
Price sourceMedian of 4 providersMedian of 9 providersMedian of 2 providers
Limits
Context window1,048,576 tokens (best)262,144 tokens262,144 tokens
Max output32,768 tokens262,144 tokens (best)32,768 tokens
Inputs and features
TextYesYesYes
ImagesNoNoYes
PDFsNoNoNo
AudioNoNoNo
VideoNoNoNo
ReasoningYesYesYes
Tool callingYesYesYes
Structured outputNoYesYes
Availability
WeightsOpenOpenOpen
API model IDpoolside/laguna-s-2.1nvidia/nemotron-3.5-lightning-30b-a3b—
API providers612 (best)2
ReleasedJul 21, 2026Aug 11, 2026Jun 9, 2026
Knowledge cutoff———
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.

  • Laguna S 2.1$1.40
  • Nemotron 3.5 Lightning 30B A3B$0.90
  • Gemma 4 12B IT$1.30
04 — Questions

Which should you choose?

Which is better: Laguna S 2.1, Nemotron 3.5 Lightning 30B A3B or Gemma 4 12B IT?

Gemma 4 12B IT is the better all-round choice, scoring 75/100 against Nemotron 3.5 Lightning 30B A3B (71) and Laguna S 2.1 (69). It leads on inputs & features. Laguna S 2.1 wins on context window. Nemotron 3.5 Lightning 30B A3B wins on 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, Laguna S 2.1, Nemotron 3.5 Lightning 30B A3B or Gemma 4 12B IT?

Nemotron 3.5 Lightning 30B A3B is cheaper at $0.05 input / $0.20 output per million tokens (median across 9 API providers; free on Nvidia). Laguna S 2.1 costs $0.10 input / $0.20 output per million tokens (median across 4 API providers; free on Poolside); Gemma 4 12B IT costs $0.075 input / $0.275 output per million tokens (median across 2 API providers). At a typical mix of three input tokens to one output token, that is $0.087 per million tokens for Nemotron 3.5 Lightning 30B A3B versus $0.125 for Laguna S 2.1 (1.4× as much) and $0.125 for Gemma 4 12B IT (1.4× as much).

Which scores higher on benchmarks?

There is no independent benchmark that covers all three models yet. Laguna S 2.1 has not been scored yet, Nemotron 3.5 Lightning 30B A3B has not been scored yet and Gemma 4 12B IT has not been scored yet.

Which is better for coding?

There are no published SWE-bench Verified results for Laguna S 2.1, Nemotron 3.5 Lightning 30B A3B and Gemma 4 12B IT 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?

Laguna S 2.1 has the largest context window at 1,048,576 tokens, against 262,144 for Nemotron 3.5 Lightning 30B A3B and 262,144 for Gemma 4 12B IT. Maximum output per response: Laguna S 2.1 up to 32,768, Nemotron 3.5 Lightning 30B A3B up to 262,144, Gemma 4 12B IT up to 32,768 tokens.

Which can read images, PDFs, audio or video?

Laguna S 2.1 accepts text; Nemotron 3.5 Lightning 30B A3B accepts text; Gemma 4 12B IT accepts text and images. Gemma 4 12B IT handles the widest range of inputs.

Are any of these open source?

Yes, all three publish their weights, so you can self-host them.

Which is newer?

Nemotron 3.5 Lightning 30B A3B is the newest, released Aug 11, 2026. Laguna S 2.1 came out Jul 21, 2026; Gemma 4 12B IT came out Jun 9, 2026.

How do you decide the winner?

Each model gets a 0–100 score on capability (50%, independent benchmark results); price (25%, blended price per million tokens (3 input : 1 output), log scale); inputs & features (15%, image, PDF, audio and video input, tool calling, structured output and reasoning); context window (10%, 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 Oct 4, 2026.

05 — Keep comparing

Related comparisons