Model comparison
Gemma 2 27B vs Kimi K3
Kimi K3 is the stronger model overall, scoring 59.5 to 29.4 on the Noometry Index. Gemma 2 27B costs 9.2× less per token, which makes it the better buy when Kimi K3's lead doesn't matter for your workload.
Last verified . 22 shared benchmarks.
GoogleSummary
- They share 22 benchmarks with published results for both. Gemma 2 27B scores higher in 0 categories and Kimi K3 in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Kimi K3 leads 74.2 to 10.7.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 1.4% for Gemma 2 27B and 97.2% for Kimi K3.
- Gemma 2 27B is cheaper at $0.65 / $0.65 per million input/output tokens, against $3 / $15 for Kimi K3.
- Kimi K3 accepts more context: 1.05M tokens versus 8K.
Side by side
| Gemma 2 27B | Kimi K3 | |
|---|---|---|
| Provider | Moonshot AI | |
| Noometry Index | 29.4 | 59.5 |
| Released | 2024-06-24 | 2026-07-16 |
| Weights | Open | Open |
| Context window | 8K | 1.05M |
| Max output | 2K | 1.05M |
| Input $ / M tokens | $0.65 | $3 |
| Output $ / M tokens | $0.65 | $15 |
| Results tracked | 34 | 53 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Kimi K3 leads
Gemma 2 27B: 34.1 (#246), Kimi K3: 61.0 (#10)
| Benchmark | Gemma 2 27B | Kimi K3 |
|---|---|---|
| LMArena Coding | 1211 | 1508 |
| DeepSWE | — | 68.5% |
| FrontierCode | — | 44.2% |
| LMArena WebDev | — | 1654 |
| FrontierSWE | — | 25.9% |
| SciCode | — | 59.5% |
| WeirdML | — | 82.6% |
| BigCodeBench Instruct | 42.8% | — |
| LiveBench Coding | 36% | — |
| BigCodeBench Complete | 52.5% | — |
| ALE-Bench | — | 1,524 |
Agentic & Tool Use Not comparable
Gemma 2 27B: —, Kimi K3: 41.8 (#20)
| Benchmark | Gemma 2 27B | 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 2 27B: 15.3 (#315), Kimi K3: 63.0 (#17)
| Benchmark | Gemma 2 27B | Kimi K3 |
|---|---|---|
| LMArena Hard Prompts | 1198 | 1496 |
| DTBench | 48% | 91.2% |
| LMCA | 7.1% | 52.7% |
| Epoch Capabilities Index | 122.08 | 157.45 |
| ARC-AGI-2 | — | 60.4% |
| SimpleBench | — | 60.7% |
| NYT Connections (extended) | — | 93.6% |
| ARC-AGI-1 | — | 94.5% |
| CritPt | — | 23.4% |
| Chess Puzzles | — | 39% |
| LiveBench Reasoning | 28.1% | — |
| Mystery Game Puzzles | — | 26% |
| LiveBench Data Analysis | 47.9% | — |
| Surface Evolver Bench | — | 95% |
| ForecastBench | — | 61.1 |
| LiveBench | 38.2% | — |
Math Kimi K3 leads
Gemma 2 27B: 10.7 (#311), Kimi K3: 74.2 (#16)
| Benchmark | Gemma 2 27B | Kimi K3 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 1.4% | 97.2% |
| LMArena Math | 1212 | 1491 |
| FrontierMath (Tiers 1-3) | — | 72.2% |
| FrontierMath Tier 4 | — | 39% |
| MathArena Final-Answer Competitions | — | 87.8% |
| ProofBench | — | 87% |
| LiveBench Math | 26.5% | — |
| MATH Level 5 | 27.9% | — |
Knowledge Kimi K3 leads
Gemma 2 27B: 19.0 (#280), Kimi K3: 63.2 (#21)
| Benchmark | Gemma 2 27B | Kimi K3 |
|---|---|---|
| GPQA Diamond | 36.5% | 93.1% |
| LMArena Expert | 1172 | 1521 |
| SimpleQA Verified | — | 50.6% |
| Confabulations | 27.1% | — |
| MMLU | 75.7% | — |
Multimodal Not comparable
Gemma 2 27B: —, Kimi K3: 37.8 (#70)
| Benchmark | Gemma 2 27B | Kimi K3 |
|---|---|---|
| Blueprint-Bench 2 | — | 29.5% |
| Furniture Assembly | — | 34.2% |
Multilingual Kimi K3 leads
Gemma 2 27B: 38.6 (#226), Kimi K3: 56.3 (#21)
| Benchmark | Gemma 2 27B | Kimi K3 |
|---|---|---|
| LMArena Non-English | 1217 | 1466 |
| LMArena Chinese | 1221 | 1529 |
| LMArena French | 1247 | 1491 |
| LMArena German | 1209 | 1488 |
| LMArena Japanese | 1175 | 1487 |
| LMArena Korean | 1174 | 1458 |
| LMArena Russian | 1234 | 1482 |
| LMArena Spanish | 1228 | 1472 |
Instruction Following Kimi K3 leads
Gemma 2 27B: 60.5 (#249), Kimi K3: 77.7 (#14)
| Benchmark | Gemma 2 27B | Kimi K3 |
|---|---|---|
| LMArena Instruction Following | 1206 | 1483 |
| LiveBench Instruction Following | 58.1% | — |
Long Context Kimi K3 leads
Gemma 2 27B: 37.3 (#218), Kimi K3: 45.8 (#29)
| Benchmark | Gemma 2 27B | Kimi K3 |
|---|---|---|
| LMArena Longer Query | 1231 | 1494 |
Writing & Preference Kimi K3 leads
Gemma 2 27B: 44.2 (#225), Kimi K3: 76.6 (#4)
| Benchmark | Gemma 2 27B | Kimi K3 |
|---|---|---|
| LMArena Text | 1231 | 1476 |
| LMArena Creative Writing | 1241 | 1454 |
| LMArena Multi-Turn | 1224 | 1488 |
| EQ-Bench Creative Writing | — | 2082 |
| EQ-Bench 4 | — | 1339 |
| LiveBench Language | 32.6% | — |
Frequently asked questions
Is Gemma 2 27B better than Kimi K3?
Kimi K3 is the stronger model overall, scoring 59.5 to 29.4 on the Noometry Index. Gemma 2 27B costs 9.2× less per token, which makes it the better buy when Kimi K3's lead doesn't matter for your workload.
Which is cheaper, Gemma 2 27B or Kimi K3?
Gemma 2 27B is cheaper. It lists at $0.65 per million input tokens and $0.65 per million output tokens; Kimi K3 lists at $3 and $15.
Is Gemma 2 27B or Kimi K3 better for coding?
Kimi K3 scores higher on coding benchmarks: 61.0 versus 34.1 in the Noometry coding category.
Which has the bigger context window?
Kimi K3 does, with 1.05M tokens against 8K.
How many benchmarks do Gemma 2 27B and Kimi K3 share?
22 benchmarks have published results for both models. Gemma 2 27B has 34 scored results on Noometry and Kimi K3 has 53.