Model comparison
Codellama 34b Instruct vs GPT-5.4
GPT-5.4 is the stronger model overall, scoring 59.4 to 30.8 on the Noometry Index.
Last verified . 10 shared benchmarks.
MetaSummary
- They share 10 benchmarks with published results for both. Codellama 34b Instruct scores higher in 0 categories and GPT-5.4 in 7 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where GPT-5.4 leads 71.9 to 28.2.
- Codellama 34b Instruct has downloadable open weights; the other is API-only.
Side by side
| Codellama 34b Instruct | GPT-5.4 | |
|---|---|---|
| Provider | Meta | OpenAI |
| Noometry Index | 30.8 | 59.4 |
| Released | — | 2026-03-05 |
| Weights | Open | Proprietary |
| Context window | — | 1.05M |
| Max output | — | 128K |
| Input $ / M tokens | — | $2.50 |
| Output $ / M tokens | — | $15 |
| Results tracked | 14 | 68 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.4 leads
Codellama 34b Instruct: 28.5 (#314), GPT-5.4: 52.6 (#33)
| Benchmark | Codellama 34b Instruct | GPT-5.4 |
|---|---|---|
| LMArena Coding | 1046 | 1497 |
| SWE-bench Verified | — | 76.9% |
| DeepSWE | — | 51.8% |
| LMArena WebDev | — | 1465 |
| SciCode | — | 56.6% |
| GSO | — | 31.4% |
| WeirdML | — | 77.7% |
| BigCodeBench Instruct | 29% | — |
| MirrorCode | — | 15.6% |
| BigCodeBench Complete | 37.1% | — |
| ALE-Bench | — | 1,607 |
| AlgoTune | — | 1.85 |
| HumanEval+ | 43.9% | — |
| MBPP+ | 56.3% | — |
Agentic & Tool Use Not comparable
Codellama 34b Instruct: —, GPT-5.4: 46.5 (#13)
| Benchmark | Codellama 34b Instruct | GPT-5.4 |
|---|---|---|
| Terminal-Bench | — | 81.8% |
| APEX-Agents | — | 52.4% |
| τ²-bench Banking | — | 39.4% |
| DeepResearch Bench | — | 35.1% |
| PostTrainBench | — | 19% |
| GBAEval | — | 45.1% |
| LMArena Search | — | 1197 |
| METR Time Horizons | — | 74.3% |
| Vending-Bench 2 | — | 6,144 |
Reasoning GPT-5.4 leads
Codellama 34b Instruct: 19.6 (#255), GPT-5.4: 61.8 (#19)
| Benchmark | Codellama 34b Instruct | GPT-5.4 |
|---|---|---|
| LMArena Hard Prompts | 1032 | 1485 |
| ARC-AGI-2 | — | 74% |
| Kagi LLM Benchmark | — | 63.8% |
| NYT Connections (extended) | — | 91.3% |
| ARC-AGI-1 | — | 93.7% |
| CritPt | — | 23.4% |
| Chess Puzzles | — | 44% |
| EnigmaEval | — | 16% |
| Thematic Generalization | — | 80% |
| EBR-Bench | — | 25.4% |
| Mystery Game Puzzles | — | 37% |
| DTBench | — | 94.4% |
| LMCA | — | 52% |
| Epoch Capabilities Index | — | 156.81 |
| ForecastBench | — | 59.5 |
Math GPT-5.4 leads
Codellama 34b Instruct: 31.0 (#230), GPT-5.4: 73.5 (#19)
| Benchmark | Codellama 34b Instruct | GPT-5.4 |
|---|---|---|
| LMArena Math | 1056 | 1488 |
| FrontierMath (Tiers 1-3) | — | 78.6% |
| FrontierMath Tier 4 | — | 49% |
| MathArena Final-Answer Competitions | — | 83.1% |
| OTIS Mock AIME 2024-2025 | — | 97.8% |
| ProofBench | — | 56% |
| FrontierMath (Feb 2025 set) | — | 47.6% |
| FrontierMath Tier 4 (v1) | — | 27.1% |
Knowledge Not comparable
Codellama 34b Instruct: —, GPT-5.4: 65.3 (#14)
| Benchmark | Codellama 34b Instruct | GPT-5.4 |
|---|---|---|
| GPQA Diamond | — | 93.3% |
| Humanity's Last Exam | — | 36.2% |
| SimpleQA Verified | — | 45.1% |
| Vectara Hallucination Rate | — | 7% |
| LMArena Expert | — | 1507 |
Multimodal Not comparable
Codellama 34b Instruct: —, GPT-5.4: 43.7 (#20)
| Benchmark | Codellama 34b Instruct | GPT-5.4 |
|---|---|---|
| LMArena Vision | — | 1303 |
| Blueprint-Bench 2 | — | 27.1% |
| Furniture Assembly | — | 37.5% |
| LMArena Document | — | 1471 |
Multilingual GPT-5.4 leads
Codellama 34b Instruct: 25.8 (#284), GPT-5.4: 56.2 (#23)
| Benchmark | Codellama 34b Instruct | GPT-5.4 |
|---|---|---|
| LMArena Non-English | 1011 | 1465 |
| LMArena Chinese | 976 | 1519 |
| LMArena French | — | 1493 |
| LMArena German | — | 1472 |
| LMArena Japanese | — | 1485 |
| LMArena Korean | — | 1448 |
| LMArena Russian | — | 1480 |
| LMArena Spanish | — | 1454 |
Instruction Following GPT-5.4 leads
Codellama 34b Instruct: 52.2 (#291), GPT-5.4: 77.1 (#27)
| Benchmark | Codellama 34b Instruct | GPT-5.4 |
|---|---|---|
| LMArena Instruction Following | 1028 | 1469 |
Long Context GPT-5.4 leads
Codellama 34b Instruct: 30.9 (#284), GPT-5.4: 50.3 (#8)
| Benchmark | Codellama 34b Instruct | GPT-5.4 |
|---|---|---|
| LMArena Longer Query | 1013 | 1473 |
| CL-bench | — | 27.9% |
| CL-bench Life | — | 21.7% |
Writing & Preference GPT-5.4 leads
Codellama 34b Instruct: 28.2 (#297), GPT-5.4: 71.9 (#17)
| Benchmark | Codellama 34b Instruct | GPT-5.4 |
|---|---|---|
| LMArena Text | 1066 | 1469 |
| LMArena Creative Writing | 1032 | 1439 |
| LMArena Multi-Turn | 1015 | 1482 |
| EQ-Bench Creative Writing | — | 1840 |
| EQ-Bench 4 | — | 1272 |
Frequently asked questions
Is Codellama 34b Instruct better than GPT-5.4?
GPT-5.4 is the stronger model overall, scoring 59.4 to 30.8 on the Noometry Index.
Is Codellama 34b Instruct or GPT-5.4 better for coding?
GPT-5.4 scores higher on coding benchmarks: 52.6 versus 28.5 in the Noometry coding category.
How many benchmarks do Codellama 34b Instruct and GPT-5.4 share?
10 benchmarks have published results for both models. Codellama 34b Instruct has 14 scored results on Noometry and GPT-5.4 has 68.