Model comparison
Grok 4.20 (Non-Reasoning) vs Kimi K3
Kimi K3 is the stronger model overall, scoring 59.5 to 48.6 on the Noometry Index. Grok 4.20 (Non-Reasoning) costs 3.8× less per token, which makes it the better buy when Kimi K3's lead doesn't matter for your workload.
Last verified . 38 shared benchmarks.
Summary
- They share 38 benchmarks with published results for both. Grok 4.20 (Non-Reasoning) scores higher in 0 categories and Kimi K3 in 10 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where Kimi K3 leads 74.2 to 48.2.
- The biggest single-benchmark swing is ProofBench: 14% for Grok 4.20 (Non-Reasoning) and 87% for Kimi K3.
- Grok 4.20 (Non-Reasoning) is cheaper at $1.25 / $2.50 per million input/output tokens, against $3 / $15 for Kimi K3.
- Kimi K3 accepts more context: 1.05M tokens versus 1M.
- Kimi K3 has downloadable open weights; the other is API-only.
Side by side
| Grok 4.20 (Non-Reasoning) | Kimi K3 | |
|---|---|---|
| Provider | xAI | Moonshot AI |
| Noometry Index | 48.6 | 59.5 |
| Released | 2026-02-17 | 2026-07-16 |
| Weights | Proprietary | Open |
| Context window | 1M | 1.05M |
| Max output | 30K | 1.05M |
| Input $ / M tokens | $1.25 | $3 |
| Output $ / M tokens | $2.50 | $15 |
| Results tracked | 46 | 53 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Kimi K3 leads
Grok 4.20 (Non-Reasoning): 42.1 (#112), Kimi K3: 61.0 (#10)
| Benchmark | Grok 4.20 (Non-Reasoning) | Kimi K3 |
|---|---|---|
| LMArena WebDev | 1375 | 1654 |
| WeirdML | 52.3% | 82.6% |
| LMArena Coding | 1459 | 1508 |
| ALE-Bench | 1,150 | 1,524 |
| DeepSWE | — | 68.5% |
| FrontierCode | — | 44.2% |
| FrontierSWE | — | 25.9% |
| SciCode | — | 59.5% |
Agentic & Tool Use Kimi K3 leads
Grok 4.20 (Non-Reasoning): 34.4 (#46), Kimi K3: 41.8 (#20)
| Benchmark | Grok 4.20 (Non-Reasoning) | Kimi K3 |
|---|---|---|
| τ²-bench Banking | 18% | 37.1% |
| Vending-Bench 2 | 4,663 | 5,165 |
| Terminal-Bench | 57.3% | — |
| APEX-Agents | — | 50.6% |
| PostTrainBench | — | 32% |
| GBAEval | — | 48.3% |
| GDP.pdf | — | 19% |
| LMArena Search | 1189 | — |
Reasoning Kimi K3 leads
Grok 4.20 (Non-Reasoning): 52.3 (#32), Kimi K3: 63.0 (#17)
| Benchmark | Grok 4.20 (Non-Reasoning) | Kimi K3 |
|---|---|---|
| ARC-AGI-2 | 65.1% | 60.4% |
| NYT Connections (extended) | 85.4% | 93.6% |
| ARC-AGI-1 | 89.5% | 94.5% |
| Chess Puzzles | 24% | 39% |
| LMArena Hard Prompts | 1451 | 1496 |
| DTBench | 90.1% | 91.2% |
| LMCA | 38.7% | 52.7% |
| Epoch Capabilities Index | 151.98 | 157.45 |
| ForecastBench | 61.4 | 61.1 |
| SimpleBench | — | 60.7% |
| Kagi LLM Benchmark | 75% | — |
| CritPt | — | 23.4% |
| Thematic Generalization | 63.8% | — |
| Mystery Game Puzzles | — | 26% |
| Surface Evolver Bench | — | 95% |
Math Kimi K3 leads
Grok 4.20 (Non-Reasoning): 48.2 (#65), Kimi K3: 74.2 (#16)
| Benchmark | Grok 4.20 (Non-Reasoning) | Kimi K3 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 44.9% | 72.2% |
| FrontierMath Tier 4 | 17.1% | 39% |
| OTIS Mock AIME 2024-2025 | 92.2% | 97.2% |
| ProofBench | 14% | 87% |
| LMArena Math | 1455 | 1491 |
| MathArena Final-Answer Competitions | — | 87.8% |
Knowledge Kimi K3 leads
Grok 4.20 (Non-Reasoning): 52.8 (#60), Kimi K3: 63.2 (#21)
| Benchmark | Grok 4.20 (Non-Reasoning) | Kimi K3 |
|---|---|---|
| GPQA Diamond | 89.3% | 93.1% |
| SimpleQA Verified | 30.2% | 50.6% |
| LMArena Expert | 1439 | 1521 |
Multimodal Kimi K3 leads
Grok 4.20 (Non-Reasoning): 33.3 (#98), Kimi K3: 37.8 (#70)
| Benchmark | Grok 4.20 (Non-Reasoning) | Kimi K3 |
|---|---|---|
| Blueprint-Bench 2 | 0% | 29.5% |
| LMArena Vision | 1263 | — |
| Furniture Assembly | — | 34.2% |
| LMArena Document | 1416 | — |
Multilingual Kimi K3 leads
Grok 4.20 (Non-Reasoning): 54.5 (#40), Kimi K3: 56.3 (#21)
| Benchmark | Grok 4.20 (Non-Reasoning) | Kimi K3 |
|---|---|---|
| LMArena Non-English | 1441 | 1466 |
| LMArena Chinese | 1481 | 1529 |
| LMArena French | 1476 | 1491 |
| LMArena German | 1465 | 1488 |
| LMArena Japanese | 1449 | 1487 |
| LMArena Korean | 1417 | 1458 |
| LMArena Russian | 1458 | 1482 |
| LMArena Spanish | 1443 | 1472 |
Instruction Following Kimi K3 leads
Grok 4.20 (Non-Reasoning): 74.8 (#83), Kimi K3: 77.7 (#14)
| Benchmark | Grok 4.20 (Non-Reasoning) | Kimi K3 |
|---|---|---|
| LMArena Instruction Following | 1420 | 1483 |
Long Context Too close to call
Grok 4.20 (Non-Reasoning): 45.5 (#34), Kimi K3: 45.8 (#29)
| Benchmark | Grok 4.20 (Non-Reasoning) | Kimi K3 |
|---|---|---|
| LMArena Longer Query | 1437 | 1494 |
| CL-bench | 22.2% | — |
| CL-bench Life | 11.9% | — |
Writing & Preference Kimi K3 leads
Grok 4.20 (Non-Reasoning): 65.7 (#44), Kimi K3: 76.6 (#4)
| Benchmark | Grok 4.20 (Non-Reasoning) | Kimi K3 |
|---|---|---|
| LMArena Text | 1451 | 1476 |
| LMArena Creative Writing | 1438 | 1454 |
| EQ-Bench Creative Writing | 1574 | 2082 |
| LMArena Multi-Turn | 1456 | 1488 |
| EQ-Bench 4 | — | 1339 |
Frequently asked questions
Is Grok 4.20 (Non-Reasoning) better than Kimi K3?
Kimi K3 is the stronger model overall, scoring 59.5 to 48.6 on the Noometry Index. Grok 4.20 (Non-Reasoning) costs 3.8× less per token, which makes it the better buy when Kimi K3's lead doesn't matter for your workload.
Which is cheaper, Grok 4.20 (Non-Reasoning) or Kimi K3?
Grok 4.20 (Non-Reasoning) is cheaper. It lists at $1.25 per million input tokens and $2.50 per million output tokens; Kimi K3 lists at $3 and $15.
Is Grok 4.20 (Non-Reasoning) or Kimi K3 better for coding?
Kimi K3 scores higher on coding benchmarks: 61.0 versus 42.1 in the Noometry coding category.
Which has the bigger context window?
Kimi K3 does, with 1.05M tokens against 1M.
How many benchmarks do Grok 4.20 (Non-Reasoning) and Kimi K3 share?
38 benchmarks have published results for both models. Grok 4.20 (Non-Reasoning) has 46 scored results on Noometry and Kimi K3 has 53.