# Codellama 70b Instruct vs DeepSeek-R1

> DeepSeek-R1 is the stronger model overall, scoring 42.0 to 33.3 on the Noometry Index.

- Canonical page: https://noometry.com/compare/codellama-70b-instruct-vs-deepseek-r1
- Last updated: 2026-10-10
- Shared benchmarks: 4

## Summary

- They share 4 benchmarks with published results for both. Codellama 70b Instruct scores higher in 1 category and DeepSeek-R1 in 4 categories; 4 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where DeepSeek-R1 leads 61.5 to 33.6.
- Codellama 70b Instruct has downloadable open weights; the other is API-only.

## Snapshot

| | Codellama 70b Instruct | DeepSeek-R1 |
|---|---|---|
| Provider | Meta | DeepSeek |
| Noometry Index | 33.3 | 42.0 |
| Rank | 236 | 110 |
| Context | — | 164K |
| Input $/M | — | $0.50 |
| Output $/M | — | $2.15 |
| Weights | Open | Proprietary |

## Coding

- Codellama 70b Instruct: 37.0 (#189)
- DeepSeek-R1: 45.9 (#63)

| Benchmark | Codellama 70b Instruct | DeepSeek-R1 |
|---|---|---|
| Aider Polyglot | — | 71.4% |
| SciCode | — | 35.7% |
| WeirdML | — | 41.6% |
| BigCodeBench Instruct | 40.7% | — |
| LiveBench Coding | — | 66.7% |
| LMArena Coding | — | 1427 |
| BigCodeBench Complete | 49.6% | — |
| ALE-Bench | — | 804.12 |
| AlgoTune | — | 1.7 |
| HumanEval+ | 65.9% | — |

## Agentic & Tool Use

- Codellama 70b Instruct: —
- DeepSeek-R1: 30.4 (#68)

| Benchmark | Codellama 70b Instruct | DeepSeek-R1 |
|---|---|---|
| DeepResearch Bench | — | 35.1% |
| BALROG | — | 34.9% |
| METR Time Horizons | — | 53.8% |

## Reasoning

- Codellama 70b Instruct: 19.2 (#241)
- DeepSeek-R1: 18.7 (#253)

| Benchmark | Codellama 70b Instruct | DeepSeek-R1 |
|---|---|---|
| LMArena Hard Prompts | 1052 | 1415 |
| ARC-AGI-2 | — | 1.3% |
| SimpleBench | — | 40.8% |
| Kagi LLM Benchmark | — | 69.4% |
| ARC-AGI-1 | — | 21.2% |
| CritPt | — | 1.1% |
| LiveBench Reasoning | — | 83.2% |
| LiveBench Data Analysis | — | 69.8% |
| Epoch Capabilities Index | — | 141.3 |
| ForecastBench | — | 60 |
| LiveBench | — | 71.6% |

## Math

- Codellama 70b Instruct: —
- DeepSeek-R1: 42.0 (#70)

| Benchmark | Codellama 70b Instruct | DeepSeek-R1 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | — | 66.4% |
| Omni-MATH | — | 42.4% |
| LiveBench Math | — | 80.7% |
| LMArena Math | — | 1401 |
| MATH Level 5 | — | 96.6% |

## Knowledge

- Codellama 70b Instruct: —
- DeepSeek-R1: 44.3 (#81)

| Benchmark | Codellama 70b Instruct | DeepSeek-R1 |
|---|---|---|
| GPQA Diamond | — | 76.3% |
| MMLU-Pro | — | 79.3% |
| Confabulations | — | 12.7% |
| Vectara Hallucination Rate | — | 11.3% |
| GPQA (HELM) | — | 66.6% |
| LMArena Expert | — | 1394 |

## Multilingual

- Codellama 70b Instruct: 24.9 (#283)
- DeepSeek-R1: 52.6 (#80)

| Benchmark | Codellama 70b Instruct | DeepSeek-R1 |
|---|---|---|
| LMArena Non-English | 992 | 1412 |
| LMArena Chinese | — | 1443 |
| LMArena French | — | 1415 |
| LMArena German | — | 1405 |
| LMArena Japanese | — | 1394 |
| LMArena Korean | — | 1359 |
| LMArena Russian | — | 1423 |
| LMArena Spanish | — | 1410 |

## Instruction Following

- Codellama 70b Instruct: 52.1 (#288)
- DeepSeek-R1: 72.2 (#138)

| Benchmark | Codellama 70b Instruct | DeepSeek-R1 |
|---|---|---|
| LMArena Instruction Following | 1024 | 1382 |
| LiveBench Instruction Following | — | 80.5% |
| IFEval | — | 78.4% |

## Long Context

- Codellama 70b Instruct: —
- DeepSeek-R1: 45.4 (#36)

| Benchmark | Codellama 70b Instruct | DeepSeek-R1 |
|---|---|---|
| Fiction.LiveBench | — | 75% |
| LMArena Longer Query | — | 1391 |

## Writing & Preference

- Codellama 70b Instruct: 33.6 (#272)
- DeepSeek-R1: 61.5 (#84)

| Benchmark | Codellama 70b Instruct | DeepSeek-R1 |
|---|---|---|
| LMArena Text | 1057 | 1428 |
| LMArena Creative Writing | — | 1406 |
| Short-Story Creative Writing | — | 83% |
| EQ-Bench Creative Writing | — | 1500 |
| WildBench | — | 82.8% |
| LMArena Multi-Turn | — | 1404 |
| LiveBench Language | — | 48.5% |

## FAQ

### Is Codellama 70b Instruct better than DeepSeek-R1?

DeepSeek-R1 is the stronger model overall, scoring 42.0 to 33.3 on the Noometry Index.

### Is Codellama 70b Instruct or DeepSeek-R1 better for coding?

DeepSeek-R1 scores higher on coding benchmarks: 45.9 versus 37.0 in the Noometry coding category.

### How many benchmarks do Codellama 70b Instruct and DeepSeek-R1 share?

4 benchmarks have published results for both models. Codellama 70b Instruct has 7 scored results on Noometry and DeepSeek-R1 has 52.