Side-by-side benchmark comparison across agentic, coding, multimodal, knowledge, reasoning, and math workflows.
1-bit Bonsai 4B
~44
0/8 categoriesKimi K2
53
Winner · 3/8 categories1-bit Bonsai 4B· Kimi K2
Pick Kimi K2 if you want the stronger benchmark profile. 1-bit Bonsai 4B only becomes the better choice if its workflow or ecosystem matters more than the raw scoreboard.
Kimi K2 is clearly ahead on the aggregate, 53 to 44. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.
Kimi K2's sharpest advantage is in knowledge, where it averages 64 against 28.7. The single biggest benchmark swing on the page is GPQA, 28.7% to 75.1%.
Kimi K2 gives you the larger context window at 128K, compared with 32K for 1-bit Bonsai 4B.
BenchLM keeps the benchmark table and the operator tradeoffs on the same page so a better score does not hide a materially slower, pricier, or smaller-context model.
Runtime metrics show N/A when BenchLM does not have a sourced snapshot for that exact model. The scoring rules and freshness policy are documented on the methodology page.
| Benchmark | 1-bit Bonsai 4B | Kimi K2 |
|---|---|---|
| Agentic | ||
| Terminal-Bench 2.0 | — | 47.1% |
| BrowseComp | — | 60.2% |
| tau2-bench | — | 66.1% |
| Coding | ||
| SWE-bench Verified | — | 65.8% |
| LiveCodeBench | — | 53.7% |
| Multimodal & Grounded | ||
| Coming soon | ||
| Reasoning | ||
| MuSR | 41.4% | — |
| hle | — | 44.9% |
| KnowledgeKimi K2 wins | ||
| GPQA | 28.7% | 75.1% |
| MMLU | — | 89.5% |
| SuperGPQA | — | 57.2% |
| MMLU-Pro | — | 81.1% |
| SimpleQA | — | 31% |
| Instruction FollowingKimi K2 wins | ||
| IFEval | 69.6% | 89.8% |
| Multilingual | ||
| sweMultilingual | — | 61.1% |
| MathematicsKimi K2 wins | ||
| MATH-500 | 65.8% | 97.4% |
| AIME 2024 | — | 69.6% |
| AIME 2025 | — | 49.5% |
| HMMT Feb 2025 | — | 38.8% |
Kimi K2 is ahead overall, 53 to 44. The biggest single separator in this matchup is GPQA, where the scores are 28.7% and 75.1%.
Kimi K2 has the edge for knowledge tasks in this comparison, averaging 64 versus 28.7. Inside this category, GPQA is the benchmark that creates the most daylight between them.
Kimi K2 has the edge for math in this comparison, averaging 67.9 versus 65.8. Inside this category, MATH-500 is the benchmark that creates the most daylight between them.
Kimi K2 has the edge for instruction following in this comparison, averaging 89.8 versus 69.6. Inside this category, IFEval is the benchmark that creates the most daylight between them.
Get notified when new models drop, benchmark scores change, or the leaderboard shifts. One email per week.
Free. No spam. Unsubscribe anytime. We only store derived location metadata for consent routing.