The AI arena is free today

Open Superagent

IBM Granite 4.2 30B vs Nemotron 3.5 Bolt (74B A9B 24GB)

Nemotron 3.5 Bolt (74B A9B 24GB) leads the LLM Stats Score 29.6 to 23.3.

IBM · NVIDIA · Updated for 2026

Which is better?

Nemotron 3.5 Bolt (74B A9B 24GB) leads the overall LLM Stats Score 29.6 to 23.3, ranking #146 overall.

In the 3 individual benchmarks reported for both models, Nemotron 3.5 Bolt (74B A9B 24GB) wins 3; this is a narrower head-to-head signal than the composite indexes.

Based on current LLM Stats indexes, shared benchmarks, pricing, and model metadata for 2026.

Choose IBM Granite 4.2 30B

  • you want predictable pricing at $0.16/M input and $0.65/M output

Choose Nemotron 3.5 Bolt (74B A9B 24GB)

  • overall performance matters — it scores 29.6 and ranks #146 on LLM Stats
  • your work emphasizes reasoning — it leads those capability indexes
  • you value its reported benchmark strengths — it wins 3 of 3 exact shared results
  • you want the most recent training data — it shipped Oct 2026

At a glance

The differences that matter most.

Core performance indexes
23.3
#197
29.6
#146
21.0
#204
28.9
#146
8.3
#202
14.4
#154
7.4
#159
11.9
#127
Cost, coverage & limits
Benchmark wins
0 of 3
3 of 3
Input price
$0.16 / M
— / M
Output price
$0.65 / M
— / M
Context window
131,072
—

Capability indexes

Additional strengths measured across groups of related public benchmarks

2 shared
Index
IBM Granite 4.2 30B
Nemotron 3.5 Bolt (74B A9B 24GB)
25.0#124
23.4#140
7.0#161
5.4#169
Conservative TrueSkill rating · higher is betterHow scores work

Individual benchmarks

18 reported for IBM Granite 4.2 30B · 10 for Nemotron 3.5 Bolt (74B A9B 24GB)

3 shared

IBM Granite 4.2 30B outperforms in 0 benchmarks, while Nemotron 3.5 Bolt (74B A9B 24GB) is better at 3 benchmarks (GPQA, SciCode, Terminal-Bench 2.1).

Nemotron 3.5 Bolt (74B A9B 24GB) significantly outperforms across most benchmarks.

Sat Oct 24 2026 • llm-stats.com

Human preference

Blind head-to-head votes and playground preference scores

Model Size

Parameter count comparison

44.0B diff

Nemotron 3.5 Bolt (74B A9B 24GB) has 44.0B more parameters than IBM Granite 4.2 30B, making it 146.7% larger.

IBM
IBM Granite 4.2 30B
30.0Bparameters
NVIDIA
Nemotron 3.5 Bolt (74B A9B 24GB)
74.0Bparameters
30.0B
IBM Granite 4.2 30B
74.0B
Nemotron 3.5 Bolt (74B A9B 24GB)

Context Window

Maximum input and output token capacity

Only IBM Granite 4.2 30B specifies input context (131,072 tokens). Only IBM Granite 4.2 30B specifies output context (131,072 tokens).

IBM
IBM Granite 4.2 30B
Input131,072 tokens
Output131,072 tokens
NVIDIA
Nemotron 3.5 Bolt (74B A9B 24GB)
Input- tokens
Output- tokens
Sat Oct 24 2026 • llm-stats.com

Input capabilities

Documented input modalities across available providers

Nemotron 3.5 Bolt (74B A9B 24GB) supports multimodal inputs, whereas IBM Granite 4.2 30B does not.

Nemotron 3.5 Bolt (74B A9B 24GB) can handle both text and other forms of data like images, making it suitable for multimodal applications.

IBM Granite 4.2 30B

Text
Images
Audio
Video

Nemotron 3.5 Bolt (74B A9B 24GB)

Text
Images
Audio
Video

License

Usage and distribution terms

IBM Granite 4.2 30B is licensed under Apache 2.0, while Nemotron 3.5 Bolt (74B A9B 24GB) uses OpenMDW License v1.1.

License differences may affect how you can use these models in commercial or open-source projects.

IBM Granite 4.2 30B

Apache 2.0

Open weights

Nemotron 3.5 Bolt (74B A9B 24GB)

OpenMDW License v1.1

Open weights

Release Timeline

When each model was launched

IBM Granite 4.2 30B was released on 2026-08-25, while Nemotron 3.5 Bolt (74B A9B 24GB) was released on 2026-10-15.

Nemotron 3.5 Bolt (74B A9B 24GB) is 2 months newer than IBM Granite 4.2 30B.

IBM Granite 4.2 30B

Aug 25, 2026

1 months ago

Nemotron 3.5 Bolt (74B A9B 24GB)

Oct 15, 2026

1 weeks ago

1mo newer

Knowledge Cutoff

When training data ends

Nemotron 3.5 Bolt (74B A9B 24GB) has a documented knowledge cutoff of 2026-05-01, while IBM Granite 4.2 30B's cutoff date is not specified.

We can confirm Nemotron 3.5 Bolt (74B A9B 24GB)'s training data extends to 2026-05-01, but cannot make a direct comparison without IBM Granite 4.2 30B's cutoff date.

IBM Granite 4.2 30B

—

Nemotron 3.5 Bolt (74B A9B 24GB)

May 2026

Outputs Comparison

Notice missing or incorrect data?

Judge for yourself.

Run your own prompts against IBM Granite 4.2 30B and Nemotron 3.5 Bolt (74B A9B 24GB) side-by-side, then vote on the output you prefer.

IBM Granite 4.2 30B
✓ Preferred
Nemotron 3.5 Bolt (74B A9B 24GB)
Open in Playground

FAQ

Common questions about IBM Granite 4.2 30B vs Nemotron 3.5 Bolt (74B A9B 24GB).

Which is better, IBM Granite 4.2 30B or Nemotron 3.5 Bolt (74B A9B 24GB)?

Nemotron 3.5 Bolt (74B A9B 24GB) leads the LLM Stats Score 29.6 to 23.3. IBM Granite 4.2 30B is made by IBM and Nemotron 3.5 Bolt (74B A9B 24GB) is made by NVIDIA. The best choice depends on your use case — compare their capability indexes, individual benchmarks, pricing, and limits above.

How does IBM Granite 4.2 30B compare to Nemotron 3.5 Bolt (74B A9B 24GB) in benchmarks?

IBM Granite 4.2 30B scores RULER 64k: 90.0%, AIME 2025: 89.2%, HMMT25: 89.2%, RULER 128k: 81.4%, MMLU-Pro: 77.6%. Nemotron 3.5 Bolt (74B A9B 24GB) scores PinchBench: 87.8%, GPQA: 86.4%, BrowseComp: 66.8%, AA-LCR: 64.1%, MCP Atlas: 52.2%.

What are the context window sizes for IBM Granite 4.2 30B and Nemotron 3.5 Bolt (74B A9B 24GB)?

IBM Granite 4.2 30B supports 131K tokens and Nemotron 3.5 Bolt (74B A9B 24GB) supports an unknown number of tokens. A larger context window lets you process longer documents, conversations, or codebases in a single request.

What are the main differences between IBM Granite 4.2 30B and Nemotron 3.5 Bolt (74B A9B 24GB)?

Key differences include LLM Stats Score (23.3 vs 29.6), multimodal support (no vs yes), licensing (Apache 2.0 vs OpenMDW License v1.1). See the full comparison above for benchmark-by-benchmark results.

Who makes IBM Granite 4.2 30B and Nemotron 3.5 Bolt (74B A9B 24GB)?

IBM Granite 4.2 30B is developed by IBM and Nemotron 3.5 Bolt (74B A9B 24GB) is developed by NVIDIA.