Model comparison
GPT-4.1 mini vs Llama-3.3-70B-Instruct
GPT-4.1 mini is the stronger model overall, scoring 33.6 to 30.6 on the Noometry Index. Llama-3.3-70B-Instruct costs 4.5× less per token, which makes it the better buy when GPT-4.1 mini's lead doesn't matter for your workload.
Last verified . 29 shared benchmarks.
MetaSummary
- They share 29 benchmarks with published results for both. GPT-4.1 mini scores higher in 7 categories and Llama-3.3-70B-Instruct in 2 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-4.1 mini leads 24.1 to 15.3.
- The biggest single-benchmark swing is MATH Level 5: 87.3% for GPT-4.1 mini and 41.6% for Llama-3.3-70B-Instruct.
- Llama-3.3-70B-Instruct is cheaper at $0.10 / $0.32 per million input/output tokens, against $0.40 / $1.60 for GPT-4.1 mini.
- GPT-4.1 mini accepts more context: 1.05M tokens versus 128K.
- Llama-3.3-70B-Instruct has downloadable open weights; the other is API-only.
Side by side
| GPT-4.1 mini | Llama-3.3-70B-Instruct | |
|---|---|---|
| Provider | OpenAI | Meta |
| Noometry Index | 33.6 | 30.6 |
| Released | 2025-04-14 | 2024-12-06 |
| Weights | Proprietary | Open |
| Context window | 1.05M | 128K |
| Max output | 33K | 4K |
| Input $ / M tokens | $0.40 | $0.10 |
| Output $ / M tokens | $1.60 | $0.32 |
| Results tracked | 47 | 43 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
GPT-4.1 mini: 30.6 (#293), Llama-3.3-70B-Instruct: 31.0 (#290)
| Benchmark | GPT-4.1 mini | Llama-3.3-70B-Instruct |
|---|---|---|
| SciCode | 40.4% | 26% |
| WeirdML | 37.6% | 14.4% |
| BigCodeBench Instruct | 48.9% | 46.9% |
| LMArena Coding | 1367 | 1268 |
| SWE-bench Verified (bash only) | 23.9% | — |
| Aider Polyglot | 32.4% | — |
| LiveBench Coding | — | 36.6% |
| BigCodeBench Complete | — | 57.5% |
| CadEval | 16% | — |
Agentic & Tool Use GPT-4.1 mini leads
GPT-4.1 mini: 33.3 (#55), Llama-3.3-70B-Instruct: 25.8 (#105)
| Benchmark | GPT-4.1 mini | Llama-3.3-70B-Instruct |
|---|---|---|
| Berkeley Function Calling Leaderboard | 50.5% | 31.9% |
| BALROG | — | 23% |
Reasoning Llama-3.3-70B-Instruct leads
GPT-4.1 mini: 10.8 (#340), Llama-3.3-70B-Instruct: 14.1 (#327)
| Benchmark | GPT-4.1 mini | Llama-3.3-70B-Instruct |
|---|---|---|
| CritPt | 0% | 0% |
| LMArena Hard Prompts | 1349 | 1257 |
| DTBench | 68.8% | 59.5% |
| LMCA | 21.1% | 17.5% |
| Epoch Capabilities Index | 135.01 | 127.33 |
| ARC-AGI-2 | 0% | — |
| SimpleBench | — | 19.9% |
| Kagi LLM Benchmark | 48.6% | — |
| ARC-AGI-1 | 3.5% | — |
| Chess Puzzles | 7% | — |
| LiveBench Reasoning | — | 50.8% |
| Mystery Game Puzzles | 7% | — |
| LiveBench Data Analysis | — | 49.5% |
| ForecastBench | — | 58.6 |
| LiveBench | — | 50.2% |
Math GPT-4.1 mini leads
GPT-4.1 mini: 24.1 (#270), Llama-3.3-70B-Instruct: 15.3 (#298)
| Benchmark | GPT-4.1 mini | Llama-3.3-70B-Instruct |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 44.7% | 5.1% |
| LMArena Math | 1343 | 1267 |
| MATH Level 5 | 87.3% | 41.6% |
| FrontierMath (Tiers 1-3) | 6.7% | — |
| Omni-MATH | 49.1% | — |
| LiveBench Math | — | 42.2% |
| FrontierMath (Feb 2025 set) | 4.5% | — |
Knowledge GPT-4.1 mini leads
GPT-4.1 mini: 34.7 (#194), Llama-3.3-70B-Instruct: 30.6 (#226)
| Benchmark | GPT-4.1 mini | Llama-3.3-70B-Instruct |
|---|---|---|
| GPQA Diamond | 65.8% | 47.4% |
| LMArena Expert | 1338 | 1225 |
| SimpleQA Verified | 12.7% | — |
| MMLU-Pro | 78.3% | — |
| Confabulations | — | 22.8% |
| Vectara Hallucination Rate | — | 4.1% |
| GPQA (HELM) | 61.4% | — |
| MMLU | — | 86.3% |
Multimodal Not comparable
GPT-4.1 mini: 35.8 (#82), Llama-3.3-70B-Instruct: —
| Benchmark | GPT-4.1 mini | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Vision | 1181 | — |
Multilingual GPT-4.1 mini leads
GPT-4.1 mini: 45.7 (#166), Llama-3.3-70B-Instruct: 39.9 (#220)
| Benchmark | GPT-4.1 mini | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Non-English | 1318 | 1236 |
| LMArena Chinese | 1329 | 1217 |
| LMArena French | 1358 | 1281 |
| LMArena German | 1351 | 1251 |
| LMArena Japanese | 1290 | 1150 |
| LMArena Korean | 1298 | 1143 |
| LMArena Russian | 1324 | 1252 |
| LMArena Spanish | 1319 | 1270 |
Instruction Following GPT-4.1 mini leads
GPT-4.1 mini: 73.7 (#118), Llama-3.3-70B-Instruct: 71.1 (#157)
| Benchmark | GPT-4.1 mini | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Instruction Following | 1333 | 1242 |
| LiveBench Instruction Following | — | 82.7% |
| IFEval | 90.4% | — |
Long Context GPT-4.1 mini leads
GPT-4.1 mini: 31.8 (#275), Llama-3.3-70B-Instruct: 26.4 (#295)
| Benchmark | GPT-4.1 mini | Llama-3.3-70B-Instruct |
|---|---|---|
| Fiction.LiveBench | 44.4% | 33.3% |
| LMArena Longer Query | 1344 | 1256 |
Writing & Preference GPT-4.1 mini leads
GPT-4.1 mini: 48.6 (#199), Llama-3.3-70B-Instruct: 47.6 (#207)
| Benchmark | GPT-4.1 mini | Llama-3.3-70B-Instruct |
|---|---|---|
| LMArena Text | 1340 | 1274 |
| LMArena Creative Writing | 1300 | 1250 |
| LMArena Multi-Turn | 1354 | 1280 |
| EQ-Bench Creative Writing | 1147 | — |
| WildBench | 83.8% | — |
| LiveBench Language | — | 39.2% |
Frequently asked questions
Is GPT-4.1 mini better than Llama-3.3-70B-Instruct?
GPT-4.1 mini is the stronger model overall, scoring 33.6 to 30.6 on the Noometry Index. Llama-3.3-70B-Instruct costs 4.5× less per token, which makes it the better buy when GPT-4.1 mini's lead doesn't matter for your workload.
Which is cheaper, GPT-4.1 mini or Llama-3.3-70B-Instruct?
Llama-3.3-70B-Instruct is cheaper. It lists at $0.10 per million input tokens and $0.32 per million output tokens; GPT-4.1 mini lists at $0.40 and $1.60.
Is GPT-4.1 mini or Llama-3.3-70B-Instruct better for coding?
They score almost the same on coding (30.6 vs 31.0); test both on your own repository before choosing.
Which has the bigger context window?
GPT-4.1 mini does, with 1.05M tokens against 128K.
How many benchmarks do GPT-4.1 mini and Llama-3.3-70B-Instruct share?
29 benchmarks have published results for both models. GPT-4.1 mini has 47 scored results on Noometry and Llama-3.3-70B-Instruct has 43.