Side-by-side benchmark comparison across agentic, coding, multimodal, knowledge, reasoning, and math workflows.
Holo3-35B-A3B
~78
Winner · 1/8 categoriesMixtral 8x22B Instruct v0.1
36
0/8 categoriesHolo3-35B-A3B· Mixtral 8x22B Instruct v0.1
Pick Holo3-35B-A3B if you want the stronger benchmark profile. Mixtral 8x22B Instruct v0.1 only becomes the better choice if you want the cheaper token bill.
Holo3-35B-A3B is clearly ahead on the aggregate, 78 to 36. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.
Holo3-35B-A3B's sharpest advantage is in agentic, where it averages 77.8 against 31.8. The single biggest benchmark swing on the page is OSWorld-Verified, 77.8% to 28%.
Holo3-35B-A3B is also the more expensive model on tokens at $0.25 input / $1.80 output per 1M tokens, versus $0.00 input / $0.00 output per 1M tokens for Mixtral 8x22B Instruct v0.1. That is roughly Infinityx on output cost alone.
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 | Holo3-35B-A3B | Mixtral 8x22B Instruct v0.1 |
|---|---|---|
| AgenticHolo3-35B-A3B wins | ||
| OSWorld-Verified | 77.8% | 28% |
| Terminal-Bench 2.0 | — | 35% |
| BrowseComp | — | 32% |
| Coding | ||
| HumanEval | — | 54.8% |
| SWE-bench Pro | — | 40% |
| Multimodal & Grounded | ||
| MMMU-Pro | — | 35% |
| OfficeQA Pro | — | 36% |
| Reasoning | ||
| LongBench v2 | — | 39% |
| MRCRv2 | — | 38% |
| Knowledge | ||
| MMLU | — | 77.8% |
| FrontierScience | — | 53% |
| Instruction Following | ||
| Coming soon | ||
| Multilingual | ||
| MMLU-ProX | — | 42% |
| Mathematics | ||
| Coming soon | ||
Holo3-35B-A3B is ahead overall, 78 to 36. The biggest single separator in this matchup is OSWorld-Verified, where the scores are 77.8% and 28%.
Holo3-35B-A3B has the edge for agentic tasks in this comparison, averaging 77.8 versus 31.8. Inside this category, OSWorld-Verified 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.