# DeepSeek LLM 67B vs Granite 4.0 Micro

> Granite 4.0 Micro is the stronger model overall, scoring 28.8 to 24.8 on the Noometry Index.

- Canonical page: https://noometry.com/compare/deepseek-llm-67b-vs-granite-4-0-micro
- Last updated: 2026-10-10
- Shared benchmarks: 3

## Summary

- They share 3 benchmarks with published results for both. DeepSeek LLM 67B scores higher in 0 categories and Granite 4.0 Micro in 5 categories; 5 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Granite 4.0 Micro leads 46.7 to 31.8.

## Snapshot

| | DeepSeek LLM 67B | Granite 4.0 Micro |
|---|---|---|
| Provider | DeepSeek | IBM |
| Noometry Index | 24.8 | 28.8 |
| Rank | 341 | 312 |
| Context | — | 131K |
| Input $/M | — | $0.017 |
| Output $/M | — | $0.11 |
| Weights | Open | Open |

## Coding

- DeepSeek LLM 67B: 31.3 (#273)
- Granite 4.0 Micro: —

| Benchmark | DeepSeek LLM 67B | Granite 4.0 Micro |
|---|---|---|
| LMArena Coding | 1096 | — |

## Reasoning

- DeepSeek LLM 67B: 15.9 (#300)
- Granite 4.0 Micro: 18.5 (#261)

| Benchmark | DeepSeek LLM 67B | Granite 4.0 Micro |
|---|---|---|
| Chess Puzzles | 0% | 0% |
| LMArena Hard Prompts | 1070 | — |
| Epoch Capabilities Index | 110.54 | — |

## Math

- DeepSeek LLM 67B: 8.1 (#318)
- Granite 4.0 Micro: 11.2 (#301)

| Benchmark | DeepSeek LLM 67B | Granite 4.0 Micro |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 0.8% | 2.8% |
| Omni-MATH | — | 20.9% |
| LMArena Math | 1108 | — |
| MATH Level 5 | 6.4% | — |

## Knowledge

- DeepSeek LLM 67B: 7.1 (#308)
- Granite 4.0 Micro: 9.9 (#299)

| Benchmark | DeepSeek LLM 67B | Granite 4.0 Micro |
|---|---|---|
| GPQA Diamond | 24.6% | 28.3% |
| MMLU-Pro | — | 39.5% |
| GPQA (HELM) | — | 30.7% |

## Multilingual

- DeepSeek LLM 67B: 29.5 (#262)
- Granite 4.0 Micro: —

| Benchmark | DeepSeek LLM 67B | Granite 4.0 Micro |
|---|---|---|
| LMArena Non-English | 1073 | — |
| LMArena Chinese | 1132 | — |

## Instruction Following

- DeepSeek LLM 67B: 55.7 (#272)
- Granite 4.0 Micro: 70.0 (#164)

| Benchmark | DeepSeek LLM 67B | Granite 4.0 Micro |
|---|---|---|
| IFEval | — | 84.9% |
| LMArena Instruction Following | 1079 | — |

## Long Context

- DeepSeek LLM 67B: 33.1 (#260)
- Granite 4.0 Micro: —

| Benchmark | DeepSeek LLM 67B | Granite 4.0 Micro |
|---|---|---|
| LMArena Longer Query | 1092 | — |

## Writing & Preference

- DeepSeek LLM 67B: 31.8 (#277)
- Granite 4.0 Micro: 46.7 (#212)

| Benchmark | DeepSeek LLM 67B | Granite 4.0 Micro |
|---|---|---|
| LMArena Text | 1105 | — |
| LMArena Creative Writing | 1067 | — |
| WildBench | — | 67% |
| LMArena Multi-Turn | 1082 | — |

## FAQ

### Is DeepSeek LLM 67B better than Granite 4.0 Micro?

Granite 4.0 Micro is the stronger model overall, scoring 28.8 to 24.8 on the Noometry Index.

### How many benchmarks do DeepSeek LLM 67B and Granite 4.0 Micro share?

3 benchmarks have published results for both models. DeepSeek LLM 67B has 15 scored results on Noometry and Granite 4.0 Micro has 8.