Model comparison
Inkling vs Llama 4 Maverick
Inkling is the stronger model overall, scoring 44.1 to 30.9 on the Noometry Index. Llama 4 Maverick costs 8.5× less per token, which makes it the better buy when Inkling's lead doesn't matter for your workload.
Last verified . 30 shared benchmarks.
MetaSummary
- They share 30 benchmarks with published results for both. Inkling scores higher in 9 categories and Llama 4 Maverick in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Inkling leads 40.4 to 10.1.
- The biggest single-benchmark swing is ARC-AGI-1: 79.5% for Inkling and 4.4% for Llama 4 Maverick.
- Llama 4 Maverick is cheaper at $0.19 / $0.65 per million input/output tokens, against $1.87 / $4.68 for Inkling.
- Llama 4 Maverick accepts more context: 128K tokens versus 66K.
Side by side
| Inkling | Llama 4 Maverick | |
|---|---|---|
| Provider | Thinking Machines Lab | Meta |
| Noometry Index | 44.1 | 30.9 |
| Released | 2026-07-15 | 2025-04-05 |
| Weights | Open | Open |
| Context window | 66K | 128K |
| Max output | 66K | 4K |
| Input $ / M tokens | $1.87 | $0.19 |
| Output $ / M tokens | $4.68 | $0.65 |
| Results tracked | 41 | 54 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Inkling leads
Inkling: 34.5 (#234), Llama 4 Maverick: 26.6 (#324)
| Benchmark | Inkling | Llama 4 Maverick |
|---|---|---|
| SciCode | 47% | 33.1% |
| WeirdML | 32.3% | 24.5% |
| LMArena Coding | 1464 | 1302 |
| ALE-Bench | 946 | 172.97 |
| FrontierCode | 14% | — |
| SWE-bench Verified (bash only) | — | 21% |
| Aider Polyglot | — | 15.6% |
| LMArena WebDev | 1413 | — |
| FrontierSWE | 4.1% | — |
| BigCodeBench Instruct | — | 49.7% |
| BigCodeBench Complete | — | 61.4% |
Agentic & Tool Use Inkling leads
Inkling: 29.6 (#85), Llama 4 Maverick: 28.2 (#91)
| Benchmark | Inkling | Llama 4 Maverick |
|---|---|---|
| APEX-Agents | 33.8% | — |
| Berkeley Function Calling Leaderboard | — | 37.3% |
| τ²-bench Banking | 25% | — |
Reasoning Inkling leads
Inkling: 40.4 (#56), Llama 4 Maverick: 10.1 (#342)
| Benchmark | Inkling | Llama 4 Maverick |
|---|---|---|
| ARC-AGI-2 | 36.5% | 0% |
| SimpleBench | 50% | 27.7% |
| ARC-AGI-1 | 79.5% | 4.4% |
| CritPt | 5.4% | 0% |
| LMArena Hard Prompts | 1451 | 1281 |
| DTBench | 87.5% | 61.9% |
| LMCA | 37.6% | 15.9% |
| Epoch Capabilities Index | 148.54 | 132.2 |
| Kagi LLM Benchmark | — | 55.9% |
| NYT Connections (extended) | — | 8% |
| Chess Puzzles | 21% | — |
| EnigmaEval | — | 0.6% |
| ForecastBench | — | 57.5 |
Math Inkling leads
Inkling: 31.3 (#225), Llama 4 Maverick: 26.0 (#262)
| Benchmark | Inkling | Llama 4 Maverick |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 88.9% | 20.6% |
| LMArena Math | 1479 | 1299 |
| FrontierMath (Tiers 1-3) | 33.3% | — |
| FrontierMath Tier 4 | 4.9% | — |
| ProofBench | 0% | — |
| Omni-MATH | — | 42.2% |
| MATH Level 5 | — | 73% |
| FrontierMath (Feb 2025 set) | — | 0.7% |
Knowledge Inkling leads
Inkling: 55.1 (#49), Llama 4 Maverick: 33.4 (#204)
| Benchmark | Inkling | Llama 4 Maverick |
|---|---|---|
| GPQA Diamond | 88.3% | 67% |
| LMArena Expert | 1465 | 1259 |
| Humanity's Last Exam | — | 5.7% |
| SimpleQA Verified | 40.3% | — |
| MMLU-Pro | — | 81% |
| Confabulations | — | 22.6% |
| Vectara Hallucination Rate | — | 8.2% |
| GPQA (HELM) | — | 65% |
Multimodal Not comparable
Inkling: —, Llama 4 Maverick: 31.6 (#105)
| Benchmark | Inkling | Llama 4 Maverick |
|---|---|---|
| LMArena Vision | — | 1142 |
| GeoBench | — | 52% |
| SpatialViz-Bench | — | 31.8% |
Multilingual Inkling leads
Inkling: 54.0 (#52), Llama 4 Maverick: 42.2 (#195)
| Benchmark | Inkling | Llama 4 Maverick |
|---|---|---|
| LMArena Non-English | 1434 | 1269 |
| LMArena Chinese | 1490 | 1277 |
| LMArena French | 1458 | 1259 |
| LMArena German | 1446 | 1291 |
| LMArena Japanese | 1429 | 1207 |
| LMArena Korean | 1404 | 1203 |
| LMArena Russian | 1429 | 1286 |
| LMArena Spanish | 1448 | 1293 |
Instruction Following Inkling leads
Inkling: 75.1 (#71), Llama 4 Maverick: 71.7 (#146)
| Benchmark | Inkling | Llama 4 Maverick |
|---|---|---|
| LMArena Instruction Following | 1426 | 1267 |
| IFEval | — | 90.8% |
Long Context Inkling leads
Inkling: 43.8 (#86), Llama 4 Maverick: 31.4 (#279)
| Benchmark | Inkling | Llama 4 Maverick |
|---|---|---|
| LMArena Longer Query | 1434 | 1280 |
| Fiction.LiveBench | — | 46.2% |
Writing & Preference Inkling leads
Inkling: 65.2 (#51), Llama 4 Maverick: 38.8 (#252)
| Benchmark | Inkling | Llama 4 Maverick |
|---|---|---|
| LMArena Text | 1441 | 1287 |
| LMArena Creative Writing | 1387 | 1267 |
| EQ-Bench Creative Writing | 1611 | 860 |
| LMArena Multi-Turn | 1436 | 1289 |
| Short-Story Creative Writing | — | 62% |
| WildBench | — | 80% |
| EQ-Bench 4 | 1226 | — |
Frequently asked questions
Is Inkling better than Llama 4 Maverick?
Inkling is the stronger model overall, scoring 44.1 to 30.9 on the Noometry Index. Llama 4 Maverick costs 8.5× less per token, which makes it the better buy when Inkling's lead doesn't matter for your workload.
Which is cheaper, Inkling or Llama 4 Maverick?
Llama 4 Maverick is cheaper. It lists at $0.19 per million input tokens and $0.65 per million output tokens; Inkling lists at $1.87 and $4.68.
Is Inkling or Llama 4 Maverick better for coding?
Inkling scores higher on coding benchmarks: 34.5 versus 26.6 in the Noometry coding category.
Which has the bigger context window?
Llama 4 Maverick does, with 128K tokens against 66K.
How many benchmarks do Inkling and Llama 4 Maverick share?
30 benchmarks have published results for both models. Inkling has 41 scored results on Noometry and Llama 4 Maverick has 54.