Nemotron 3.5 Lightning 30B A3B vs Gemma 4 12B IT vs Granite 4.2 8B
No independent capability benchmark covers every model here yet. The score uses price, inputs & features and context window.
Price in rating: Included · Exclude price
NVIDIA
Nemotron 3.5 Lightning 30B A3B
69/100- ECI—
- Price$0.06 / $0.20
- Context262K
Google
Gemma 4 12B IT
73/100- ECI—
- Price$0.05 / $0.25
- Context262K
IBM
Granite 4.2 8B
65/100- ECI—
- Price$0.06 / $0.15
- Context131K
Capability evidence is incomplete
There is not enough shared capability evidence to name an overall winner. The scores (Gemma 4 12B IT 73/100, Nemotron 3.5 Lightning 30B A3B 69/100, Granite 4.2 8B 65/100) 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 · Gemma 4 12B IT $0.10 per 1M tokens (3:1 blend)
- Longest contextNemotron 3.5 Lightning 30B A3B and Gemma 4 12B ITNemotron 3.5 Lightning 30B A3B 262,144 · Gemma 4 12B IT 262,144 · Granite 4.2 8B 131,072 tokens
- Widest inputsGemma 4 12B ITNemotron 3.5 Lightning 30B A3B: Text · Gemma 4 12B IT: Text, Images · Granite 4.2 8B: Text
- Self-hostingAll open weightsEvery model here can be downloaded and run on your own hardware
| Measure | Weight | Nemotron 3.5 Lightning 30B A3B | Gemma 4 12B IT | Granite 4.2 8B |
|---|---|---|---|---|
| Price | 50% | 94 | 93 | 97 |
| Inputs & features | 30% | 45 | 60 | 35 |
| Context window | 20% | 43 | 43 | 32 |
| Specs & price only | 100% | 69/100 | 73/100 | 65/100 |
Left out because at least one model lacks the data: capability. The remaining weights were rescaled.
Every spec in one table
Highlighted cells lead their row. Dashes mean the data is not published.
| Specification | |||
|---|---|---|---|
| Capability | |||
| Capabilities Index (ECI) | — | — | — |
| ECI rank | — | — | — |
| Price per million tokens | |||
| Input | $0.06 | $0.05 (best) | $0.06 |
| Output | $0.20 | $0.25 | $0.15 (best) |
| Cached input | — | — | — |
| Blended (3:1) | $0.095 | $0.10 | $0.083 (best) |
| Long-context rate | Same rate | Same rate | Same rate |
| Price source | Median of 9 providers | Median of 3 providers | Median of 7 providers |
| Limits | |||
| Context window | 262,144 tokens (best) | 262,144 tokens (best) | 131,072 tokens |
| Max output | 262,144 tokens (best) | 32,768 tokens | 117,964 tokens |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | No | Yes | No |
| PDFs | No | No | No |
| Audio | No | No | No |
| Video | No | No | No |
| Reasoning | Yes | Yes | Yes |
| Tool calling | Yes | Yes | Yes |
| Structured output | Yes | No | No |
| Availability | |||
| Weights | Open | Open | Open |
| API model ID | nvidia/nemotron-3.5-lightning-30b-a3b | — | — |
| API providers | 15 (best) | 3 | 7 |
| Released | Aug 11, 2026 | Jun 9, 2026 | Aug 25, 2026 |
| Knowledge cutoff | Sep 2025 | — | — |
What would a month cost?
Enter your expected volume in millions of tokens. List prices only; caching and batch discounts would lower these.
Nemotron 3.5 Lightning 30B A3B$1.00
Gemma 4 12B IT$1.00
Granite 4.2 8B$0.90
Which should you choose?
Which is better: Nemotron 3.5 Lightning 30B A3B, Gemma 4 12B IT or Granite 4.2 8B?
There is not enough shared capability evidence to name an overall winner. The scores (Gemma 4 12B IT 73/100, Nemotron 3.5 Lightning 30B A3B 69/100, Granite 4.2 8B 65/100) 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, Nemotron 3.5 Lightning 30B A3B, Gemma 4 12B IT or Granite 4.2 8B?
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); Gemma 4 12B IT costs $0.05 input / $0.25 output per million tokens (median across 3 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 $0.10 for Gemma 4 12B IT (1.2× as much).
Which scores higher on benchmarks?
There is no independent benchmark that covers all three models yet. Nemotron 3.5 Lightning 30B A3B has not been scored yet, Gemma 4 12B IT has not been scored yet and Granite 4.2 8B has not been scored yet.
Which is better for coding?
There are no published SWE-bench Verified results for Nemotron 3.5 Lightning 30B A3B, Gemma 4 12B IT and Granite 4.2 8B 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 and Gemma 4 12B IT have the largest context windows (262,144 and 262,144 tokens), against 131,072 for Granite 4.2 8B. Maximum output per response: Nemotron 3.5 Lightning 30B A3B up to 262,144, Gemma 4 12B IT up to 32,768, Granite 4.2 8B up to 117,964 tokens.
Which can read images, PDFs, audio or video?
Nemotron 3.5 Lightning 30B A3B accepts text; Gemma 4 12B IT accepts text and images; Granite 4.2 8B accepts text. 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?
Granite 4.2 8B is the newest, released Aug 25, 2026. Nemotron 3.5 Lightning 30B A3B came out Aug 11, 2026; Gemma 4 12B IT came out Jun 9, 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.