Model comparison
GPT-4.1 vs Granite 4.0 Micro
GPT-4.1 is the stronger model overall, scoring 35.9 to 29.0 on the Noometry Index. Granite 4.0 Micro costs 86× less per token, which makes it the better buy when GPT-4.1's lead doesn't matter for your workload.
Last verified . 8 shared benchmarks.
Summary
- They share 8 benchmarks with published results for both. GPT-4.1 scores higher in 4 categories and Granite 4.0 Micro in 1 category; 5 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where GPT-4.1 leads 37.1 to 9.9.
- The biggest single-benchmark swing is MMLU-Pro: 81.1% for GPT-4.1 and 39.5% for Granite 4.0 Micro.
- Granite 4.0 Micro is cheaper at $0.017 / $0.11 per million input/output tokens, against $2 / $8 for GPT-4.1.
- GPT-4.1 accepts more context: 1.05M tokens versus 131K.
- Granite 4.0 Micro has downloadable open weights; the other is API-only.
Side by side
| GPT-4.1 | Granite 4.0 Micro | |
|---|---|---|
| Provider | OpenAI | IBM |
| Noometry Index | 35.9 | 29.0 |
| Released | 2025-04-14 | 2025-10-02 |
| Weights | Proprietary | Open |
| Context window | 1.05M | 131K |
| Max output | 33K | 118K |
| Input $ / M tokens | $2 | $0.017 |
| Output $ / M tokens | $8 | $0.11 |
| Results tracked | 52 | 8 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Not comparable
GPT-4.1: 34.4 (#238), Granite 4.0 Micro: —
| Benchmark | GPT-4.1 | Granite 4.0 Micro |
|---|---|---|
| SWE-bench Verified | 48.5% | — |
| SWE-bench Verified (bash only) | 39.6% | — |
| Aider Polyglot | 52.4% | — |
| WeirdML | 39% | — |
| LMArena Coding | 1391 | — |
| CadEval | 42% | — |
| ALE-Bench | 558.1 | — |
Agentic & Tool Use Not comparable
GPT-4.1: 34.7 (#43), Granite 4.0 Micro: —
| Benchmark | GPT-4.1 | Granite 4.0 Micro |
|---|---|---|
| Berkeley Function Calling Leaderboard | 54% | — |
Reasoning Granite 4.0 Micro leads
GPT-4.1: 11.7 (#339), Granite 4.0 Micro: 19.2 (#265)
| Benchmark | GPT-4.1 | Granite 4.0 Micro |
|---|---|---|
| Chess Puzzles | 6% | 0% |
| ARC-AGI-2 | 0.4% | — |
| SimpleBench | 27% | — |
| Kagi LLM Benchmark | 52.3% | — |
| ARC-AGI-1 | 5.5% | — |
| EnigmaEval | 2.2% | — |
| LMArena Hard Prompts | 1384 | — |
| DTBench | 68.3% | — |
| LMCA | 25.6% | — |
| Epoch Capabilities Index | 136.78 | — |
| ForecastBench | 61.5 | — |
Math GPT-4.1 leads
GPT-4.1: 22.3 (#280), Granite 4.0 Micro: 12.0 (#307)
| Benchmark | GPT-4.1 | Granite 4.0 Micro |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 38.3% | 2.8% |
| Omni-MATH | 47.1% | 20.9% |
| FrontierMath (Tiers 1-3) | 6% | — |
| LMArena Math | 1370 | — |
| MATH Level 5 | 83% | — |
| FrontierMath (Feb 2025 set) | 5.5% | — |
| FrontierMath Tier 4 (v1) | 0% | — |
Knowledge GPT-4.1 leads
GPT-4.1: 37.1 (#160), Granite 4.0 Micro: 9.9 (#304)
| Benchmark | GPT-4.1 | Granite 4.0 Micro |
|---|---|---|
| GPQA Diamond | 66.9% | 28.3% |
| MMLU-Pro | 81.1% | 39.5% |
| GPQA (HELM) | 65.9% | 30.7% |
| Humanity's Last Exam | 5.4% | — |
| SimpleQA Verified | 31.1% | — |
| Vectara Hallucination Rate | 5.6% | — |
| LMArena Expert | 1364 | — |
Multimodal Not comparable
GPT-4.1: 38.2 (#67), Granite 4.0 Micro: —
| Benchmark | GPT-4.1 | Granite 4.0 Micro |
|---|---|---|
| LMArena Vision | 1211 | — |
| GeoBench | 72% | — |
Multilingual Not comparable
GPT-4.1: 49.4 (#133), Granite 4.0 Micro: —
| Benchmark | GPT-4.1 | Granite 4.0 Micro |
|---|---|---|
| LMArena Non-English | 1370 | — |
| LMArena Chinese | 1382 | — |
| LMArena French | 1382 | — |
| LMArena German | 1381 | — |
| LMArena Japanese | 1319 | — |
| LMArena Korean | 1339 | — |
| LMArena Russian | 1377 | — |
| LMArena Spanish | 1376 | — |
Instruction Following GPT-4.1 leads
GPT-4.1: 71.3 (#153), Granite 4.0 Micro: 69.9 (#169)
| Benchmark | GPT-4.1 | Granite 4.0 Micro |
|---|---|---|
| IFEval | 83.8% | 84.9% |
| LMArena Instruction Following | 1367 | — |
Long Context Not comparable
GPT-4.1: 40.0 (#163), Granite 4.0 Micro: —
| Benchmark | GPT-4.1 | Granite 4.0 Micro |
|---|---|---|
| Fiction.LiveBench | 63.9% | — |
| LMArena Longer Query | 1385 | — |
Writing & Preference GPT-4.1 leads
GPT-4.1: 57.6 (#125), Granite 4.0 Micro: 46.7 (#216)
| Benchmark | GPT-4.1 | Granite 4.0 Micro |
|---|---|---|
| WildBench | 85.4% | 67% |
| LMArena Text | 1383 | — |
| LMArena Creative Writing | 1363 | — |
| EQ-Bench Creative Writing | 1420 | — |
| LMArena Multi-Turn | 1398 | — |
Frequently asked questions
Is GPT-4.1 better than Granite 4.0 Micro?
GPT-4.1 is the stronger model overall, scoring 35.9 to 29.0 on the Noometry Index. Granite 4.0 Micro costs 86× less per token, which makes it the better buy when GPT-4.1's lead doesn't matter for your workload.
Which is cheaper, GPT-4.1 or Granite 4.0 Micro?
Granite 4.0 Micro is cheaper. It lists at $0.017 per million input tokens and $0.11 per million output tokens; GPT-4.1 lists at $2 and $8.
Which has the bigger context window?
GPT-4.1 does, with 1.05M tokens against 131K.
How many benchmarks do GPT-4.1 and Granite 4.0 Micro share?
8 benchmarks have published results for both models. GPT-4.1 has 52 scored results on Noometry and Granite 4.0 Micro has 8.