Model comparison
GPT-5.4 mini vs Llama 3.1-8B
GPT-5.4 mini is the stronger model overall, scoring 44.8 to 22.8 on the Noometry Index. Llama 3.1-8B costs 29× less per token, which makes it the better buy when GPT-5.4 mini's lead doesn't matter for your workload.
Last verified . 27 shared benchmarks.
MetaSummary
- They share 27 benchmarks with published results for both. GPT-5.4 mini scores higher in 9 categories and Llama 3.1-8B in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where GPT-5.4 mini leads 51.4 to 8.0.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 88.9% for GPT-5.4 mini and 1.7% for Llama 3.1-8B.
- Llama 3.1-8B is cheaper at $0.05 / $0.08 per million input/output tokens, against $0.75 / $4.50 for GPT-5.4 mini.
- GPT-5.4 mini accepts more context: 400K tokens versus 128K.
- Llama 3.1-8B has downloadable open weights; the other is API-only.
Side by side
| GPT-5.4 mini | Llama 3.1-8B | |
|---|---|---|
| Provider | OpenAI | Meta |
| Noometry Index | 44.8 | 22.8 |
| Released | 2026-03-17 | 2024-07-23 |
| Weights | Proprietary | Open |
| Context window | 400K | 128K |
| Max output | 128K | 4K |
| Input $ / M tokens | $0.75 | $0.05 |
| Output $ / M tokens | $4.50 | $0.08 |
| Results tracked | 46 | 43 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.4 mini leads
GPT-5.4 mini: 45.0 (#67), Llama 3.1-8B: 19.9 (#335)
| Benchmark | GPT-5.4 mini | Llama 3.1-8B |
|---|---|---|
| SciCode | 49.9% | 13.2% |
| WeirdML | 60.3% | 1.7% |
| LMArena Coding | 1438 | 1195 |
| FrontierCode | 27% | — |
| LMArena WebDev | 1397 | — |
| BigCodeBench Instruct | — | 32.8% |
| BigCodeBench Complete | — | 40.5% |
| ALE-Bench | 1,189 | — |
| HumanEval+ | — | 62.8% |
| MBPP+ | — | 55.6% |
Agentic & Tool Use GPT-5.4 mini leads
GPT-5.4 mini: 29.6 (#74), Llama 3.1-8B: 22.3 (#123)
| Benchmark | GPT-5.4 mini | Llama 3.1-8B |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 25.8% |
| DeepResearch Bench | 36.3% | — |
| BALROG | — | 15.1% |
Reasoning GPT-5.4 mini leads
GPT-5.4 mini: 30.3 (#77), Llama 3.1-8B: 14.5 (#316)
| Benchmark | GPT-5.4 mini | Llama 3.1-8B |
|---|---|---|
| CritPt | 10% | 0% |
| Chess Puzzles | 24% | 0% |
| LMArena Hard Prompts | 1424 | 1175 |
| DTBench | 80% | 50.9% |
| LMCA | 40.8% | 5.4% |
| Epoch Capabilities Index | 148.83 | 116.48 |
| ARC-AGI-2 | 18.9% | — |
| Kagi LLM Benchmark | 37.9% | — |
| NYT Connections (extended) | 61.8% | — |
| ARC-AGI-1 | 63.7% | — |
| Thematic Generalization | 61.7% | — |
| Mystery Game Puzzles | 11% | — |
| ForecastBench | 57 | — |
| PIQA | — | 81.2% |
Math GPT-5.4 mini leads
GPT-5.4 mini: 45.4 (#63), Llama 3.1-8B: 9.5 (#311)
| Benchmark | GPT-5.4 mini | Llama 3.1-8B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 88.9% | 1.7% |
| LMArena Math | 1418 | 1179 |
| FrontierMath (Tiers 1-3) | 51.2% | — |
| FrontierMath Tier 4 | 9.8% | — |
| ProofBench | 21% | — |
| Omni-MATH | — | 13.7% |
| MATH Level 5 | — | 22.9% |
| FrontierMath (Feb 2025 set) | 28.3% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |
| GSM8K | — | 82.4% |
Knowledge GPT-5.4 mini leads
GPT-5.4 mini: 51.4 (#62), Llama 3.1-8B: 8.0 (#302)
| Benchmark | GPT-5.4 mini | Llama 3.1-8B |
|---|---|---|
| GPQA Diamond | 86.9% | 27% |
| LMArena Expert | 1434 | 1144 |
| SimpleQA Verified | 29.4% | — |
| MMLU-Pro | — | 40.6% |
| Vectara Hallucination Rate | 5.5% | — |
| GPQA (HELM) | — | 24.7% |
| BoolQ | — | 82.8% |
| MMLU | — | 56.1% |
Multimodal Not comparable
GPT-5.4 mini: 39.0 (#51), Llama 3.1-8B: —
| Benchmark | GPT-5.4 mini | Llama 3.1-8B |
|---|---|---|
| LMArena Vision | 1245 | — |
Multilingual GPT-5.4 mini leads
GPT-5.4 mini: 52.1 (#94), Llama 3.1-8B: 34.2 (#244)
| Benchmark | GPT-5.4 mini | Llama 3.1-8B |
|---|---|---|
| LMArena Non-English | 1405 | 1148 |
| LMArena Chinese | 1446 | 1151 |
| LMArena French | 1440 | 1176 |
| LMArena German | 1407 | 1144 |
| LMArena Japanese | 1375 | 1061 |
| LMArena Korean | 1369 | 1053 |
| LMArena Russian | 1417 | 1158 |
| LMArena Spanish | 1403 | 1169 |
Instruction Following GPT-5.4 mini leads
GPT-5.4 mini: 74.3 (#99), Llama 3.1-8B: 59.1 (#253)
| Benchmark | GPT-5.4 mini | Llama 3.1-8B |
|---|---|---|
| LMArena Instruction Following | 1405 | 1159 |
| IFEval | — | 74.3% |
Long Context GPT-5.4 mini leads
GPT-5.4 mini: 42.9 (#107), Llama 3.1-8B: 35.8 (#233)
| Benchmark | GPT-5.4 mini | Llama 3.1-8B |
|---|---|---|
| LMArena Longer Query | 1407 | 1182 |
Writing & Preference GPT-5.4 mini leads
GPT-5.4 mini: 64.2 (#56), Llama 3.1-8B: 29.9 (#285)
| Benchmark | GPT-5.4 mini | Llama 3.1-8B |
|---|---|---|
| LMArena Text | 1411 | 1187 |
| LMArena Creative Writing | 1370 | 1154 |
| EQ-Bench Creative Writing | 1665 | 713 |
| LMArena Multi-Turn | 1429 | 1172 |
| WildBench | — | 68.7% |
Frequently asked questions
Is GPT-5.4 mini better than Llama 3.1-8B?
GPT-5.4 mini is the stronger model overall, scoring 44.8 to 22.8 on the Noometry Index. Llama 3.1-8B costs 29× less per token, which makes it the better buy when GPT-5.4 mini's lead doesn't matter for your workload.
Which is cheaper, GPT-5.4 mini or Llama 3.1-8B?
Llama 3.1-8B is cheaper. It lists at $0.05 per million input tokens and $0.08 per million output tokens; GPT-5.4 mini lists at $0.75 and $4.50.
Is GPT-5.4 mini or Llama 3.1-8B better for coding?
GPT-5.4 mini scores higher on coding benchmarks: 45.0 versus 19.9 in the Noometry coding category.
Which has the bigger context window?
GPT-5.4 mini does, with 400K tokens against 128K.
How many benchmarks do GPT-5.4 mini and Llama 3.1-8B share?
27 benchmarks have published results for both models. GPT-5.4 mini has 46 scored results on Noometry and Llama 3.1-8B has 43.