Model comparison
Gemma 4 26B A4B IT vs Kimi K3
Kimi K3 is the stronger model overall, scoring 59.5 to 43.1 on the Noometry Index. Gemma 4 26B A4B IT costs 50× less per token, which makes it the better buy when Kimi K3's lead doesn't matter for your workload.
Last verified . 26 shared benchmarks.
GoogleSummary
- They share 26 benchmarks with published results for both. Gemma 4 26B A4B IT scores higher in 1 category and Kimi K3 in 8 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Kimi K3 leads 63.0 to 21.1.
- The biggest single-benchmark swing is WeirdML: 35.2% for Gemma 4 26B A4B IT and 82.6% for Kimi K3.
- Gemma 4 26B A4B IT is cheaper at $0.0765 / $0.26 per million input/output tokens, against $3 / $15 for Kimi K3.
- Kimi K3 accepts more context: 1.05M tokens versus 262K.
Side by side
| Gemma 4 26B A4B IT | Kimi K3 | |
|---|---|---|
| Provider | Moonshot AI | |
| Noometry Index | 43.1 | 59.5 |
| Released | 2026-04-02 | 2026-07-16 |
| Weights | Open | Open |
| Context window | 262K | 1.05M |
| Max output | 33K | 1.05M |
| Input $ / M tokens | $0.0765 | $3 |
| Output $ / M tokens | $0.26 | $15 |
| Results tracked | 28 | 53 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Kimi K3 leads
Gemma 4 26B A4B IT: 38.8 (#156), Kimi K3: 60.6 (#8)
| Benchmark | Gemma 4 26B A4B IT | Kimi K3 |
|---|---|---|
| LMArena WebDev | 1359 | 1659 |
| SciCode | 40% | 59.5% |
| WeirdML | 35.2% | 82.6% |
| LMArena Coding | 1446 | 1509 |
| ALE-Bench | 927.17 | 1,524 |
| DeepSWE | — | 68.5% |
| FrontierCode | — | 44.2% |
| FrontierSWE | — | 25.9% |
Agentic & Tool Use Not comparable
Gemma 4 26B A4B IT: —, Kimi K3: 41.8 (#19)
| Benchmark | Gemma 4 26B A4B IT | Kimi K3 |
|---|---|---|
| APEX-Agents | — | 50.6% |
| τ²-bench Banking | — | 37.1% |
| PostTrainBench | — | 32% |
| GBAEval | — | 48.3% |
| GDP.pdf | — | 19% |
| Vending-Bench 2 | — | 5,165 |
Reasoning Kimi K3 leads
Gemma 4 26B A4B IT: 21.1 (#205), Kimi K3: 63.0 (#15)
| Benchmark | Gemma 4 26B A4B IT | Kimi K3 |
|---|---|---|
| CritPt | 0% | 23.4% |
| Chess Puzzles | 6% | 39% |
| LMArena Hard Prompts | 1438 | 1495 |
| DTBench | 74.9% | 91.2% |
| LMCA | 29.7% | 52.7% |
| Epoch Capabilities Index | 141.85 | 157.68 |
| ARC-AGI-2 | — | 60.4% |
| SimpleBench | — | 60.7% |
| NYT Connections (extended) | — | 93.6% |
| ARC-AGI-1 | — | 94.5% |
| Mystery Game Puzzles | — | 26% |
| Surface Evolver Bench | — | 95% |
| ForecastBench | — | 61.1 |
Math Kimi K3 leads
Gemma 4 26B A4B IT: 45.7 (#61), Kimi K3: 74.0 (#13)
| Benchmark | Gemma 4 26B A4B IT | Kimi K3 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 82.2% | 97.2% |
| LMArena Math | 1469 | 1493 |
| FrontierMath (Tiers 1-3) | — | 72.2% |
| FrontierMath Tier 4 | — | 39% |
| MathArena Final-Answer Competitions | — | 87.8% |
| ProofBench | — | 87% |
Knowledge Kimi K3 leads
Gemma 4 26B A4B IT: 45.6 (#79), Kimi K3: 63.2 (#19)
| Benchmark | Gemma 4 26B A4B IT | Kimi K3 |
|---|---|---|
| GPQA Diamond | 73.2% | 93.1% |
| LMArena Expert | 1447 | 1520 |
| SimpleQA Verified | — | 50.6% |
| Vectara Hallucination Rate | 5.2% | — |
Multimodal Gemma 4 26B A4B IT leads
Gemma 4 26B A4B IT: 39.9 (#43), Kimi K3: 38.4 (#58)
| Benchmark | Gemma 4 26B A4B IT | Kimi K3 |
|---|---|---|
| LMArena Vision | 1260 | — |
| Blueprint-Bench 2 | — | 29.5% |
| Furniture Assembly | — | 34.2% |
Multilingual Kimi K3 leads
Gemma 4 26B A4B IT: 53.3 (#68), Kimi K3: 56.5 (#20)
| Benchmark | Gemma 4 26B A4B IT | Kimi K3 |
|---|---|---|
| LMArena Non-English | 1422 | 1466 |
| LMArena Chinese | 1495 | 1528 |
| LMArena French | 1459 | 1488 |
| LMArena Russian | 1434 | 1481 |
| LMArena Spanish | 1415 | 1478 |
| LMArena German | — | 1474 |
| LMArena Japanese | — | 1492 |
| LMArena Korean | — | 1465 |
Instruction Following Kimi K3 leads
Gemma 4 26B A4B IT: 75.0 (#78), Kimi K3: 77.8 (#13)
| Benchmark | Gemma 4 26B A4B IT | Kimi K3 |
|---|---|---|
| LMArena Instruction Following | 1420 | 1481 |
Long Context Kimi K3 leads
Gemma 4 26B A4B IT: 43.6 (#87), Kimi K3: 45.8 (#27)
| Benchmark | Gemma 4 26B A4B IT | Kimi K3 |
|---|---|---|
| LMArena Longer Query | 1427 | 1494 |
Writing & Preference Kimi K3 leads
Gemma 4 26B A4B IT: 58.8 (#110), Kimi K3: 76.7 (#4)
| Benchmark | Gemma 4 26B A4B IT | Kimi K3 |
|---|---|---|
| LMArena Text | 1434 | 1476 |
| LMArena Creative Writing | 1401 | 1452 |
| EQ-Bench Creative Writing | 1305 | 2082 |
| LMArena Multi-Turn | 1442 | 1488 |
| EQ-Bench 4 | — | 1339 |
Frequently asked questions
Is Gemma 4 26B A4B IT better than Kimi K3?
Kimi K3 is the stronger model overall, scoring 59.5 to 43.1 on the Noometry Index. Gemma 4 26B A4B IT costs 50× less per token, which makes it the better buy when Kimi K3's lead doesn't matter for your workload.
Which is cheaper, Gemma 4 26B A4B IT or Kimi K3?
Gemma 4 26B A4B IT is cheaper. It lists at $0.0765 per million input tokens and $0.26 per million output tokens; Kimi K3 lists at $3 and $15.
Is Gemma 4 26B A4B IT or Kimi K3 better for coding?
Kimi K3 scores higher on coding benchmarks: 60.6 versus 38.8 in the Noometry coding category.
Which has the bigger context window?
Kimi K3 does, with 1.05M tokens against 262K.
How many benchmarks do Gemma 4 26B A4B IT and Kimi K3 share?
26 benchmarks have published results for both models. Gemma 4 26B A4B IT has 28 scored results on Noometry and Kimi K3 has 53.