Side-by-side benchmark comparison across agentic, coding, multimodal, knowledge, reasoning, and math workflows.
1-bit Bonsai 4B
~44
1/8 categoriesZ-1
44
3/8 categories1-bit Bonsai 4B· Z-1
Treat this as a split decision. 1-bit Bonsai 4B makes more sense if mathematics is the priority; Z-1 is the better fit if knowledge is the priority or you need the larger 128K context window.
1-bit Bonsai 4B and Z-1 finish on the same overall score, so this is less about a single winner and more about where the edge shows up. The headline says tie; the benchmark table is where the real choice happens.
Z-1 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 | Z-1 |
|---|---|---|
| Agentic | ||
| Terminal-Bench 2.0 | — | 39% |
| BrowseComp | — | 49% |
| OSWorld-Verified | — | 41% |
| Coding | ||
| HumanEval | — | 44% |
| SWE-bench Verified | — | 33% |
| LiveCodeBench | — | 22% |
| SWE-bench Pro | — | 30% |
| Multimodal & Grounded | ||
| MMMU-Pro | — | 46% |
| OfficeQA Pro | — | 56% |
| ReasoningZ-1 wins | ||
| MuSR | 41.4% | 48% |
| BBH | — | 74% |
| LongBench v2 | — | 56% |
| MRCRv2 | — | 57% |
| KnowledgeZ-1 wins | ||
| GPQA | 28.7% | 51% |
| MMLU | — | 52% |
| SuperGPQA | — | 49% |
| MMLU-Pro | — | 64% |
| HLE | — | 6% |
| FrontierScience | — | 51% |
| SimpleQA | — | 50% |
| Instruction FollowingZ-1 wins | ||
| IFEval | 69.6% | 80% |
| Multilingual | ||
| MGSM | — | 74% |
| MMLU-ProX | — | 72% |
| Mathematics1-bit Bonsai 4B wins | ||
| MATH-500 | 65.8% | 73% |
| AIME 2023 | — | 52% |
| AIME 2024 | — | 54% |
| AIME 2025 | — | 53% |
| HMMT Feb 2023 | — | 48% |
| HMMT Feb 2024 | — | 50% |
| HMMT Feb 2025 | — | 49% |
| BRUMO 2025 | — | 51% |
1-bit Bonsai 4B and Z-1 are tied on overall score, so the right pick depends on which category matters most for your use case.
Z-1 has the edge for knowledge tasks in this comparison, averaging 43.1 versus 28.7. Inside this category, GPQA is the benchmark that creates the most daylight between them.
1-bit Bonsai 4B has the edge for math in this comparison, averaging 65.8 versus 57.3. Inside this category, MATH-500 is the benchmark that creates the most daylight between them.
Z-1 has the edge for reasoning in this comparison, averaging 54.2 versus 41.4. Inside this category, MuSR is the benchmark that creates the most daylight between them.
Z-1 has the edge for instruction following in this comparison, averaging 80 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.