# WizardLM-2 8x22B

> Source: Benchize (https://benchize.com) — https://benchize.com/models/wizardlm-2-8x22b. When citing or reusing this content, credit "Benchize" and link to https://benchize.com/models/wizardlm-2-8x22b.

Microsoft

Canonical URL: https://benchize.com/models/wizardlm-2-8x22b
Last updated: Thu, 22 Oct 2026 07:10:12 GMT

Reproduction and quotation must name Benchize, link to the canonical HTML URL, and retain individual source citations.

Microsoft

## Benchmarks

| Benchmark | Score | ± | Rank | Source |  |
| --- | --- | --- | --- | --- | --- |
| [SVGBench Arena](https://benchize.com/benchmarks/svgbench%3Aarena.md) | 1,468 | — | 82 / 191 | [svgbench](https://svgbench.ai/api/leaderboard?shape=2) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `rating` \| 1468 \| rating \| reported \| Higher is better \| ### Configuration Reasoning effort unknown sourceModelId model_microsoft_wizardlm_2_8x22b ### Reported statistics coverageCount 26 draws 5 judged 134 losses 58 rank 110 ratingDeviation 28 wins 71 ### Source [Original source](https://svgbench.ai/api/leaderboard?shape=2) Source svgbench Original model label Wizardlm 2 8x22b Observed 2026-10-22T04:30:04.538Z Evidence SHA-256 `5c0a9c869593820d57d2c05f6d416e1b44a740b4835e66c66e9ad8b8edc59e2c` Source locator `/rows/106` |  |  |  |  |  |
| [GPQA Diamond](https://benchize.com/benchmarks/gpqa-diamond.md) | 43.43% | SE ±2.32 pp | 118 / 126 | [epoch ai · GPQA diamond](https://epoch.ai/data/benchmark_data.zip#file=gpqa_diamond.csv) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 0.43434343434343436 \| score_fraction \| Best score (across scorers) \| Higher is better \| \| `mean_score` \| 0.43434343434343436 \| score_fraction \| mean_score \| Higher is better \| ### Configuration benchmarkVersion GPQA diamond reasoningBudget Not reported Reasoning effort Not reported sourceModelVersion WizardLM-2-8x22B ### Reported statistics Uncertainty Standard error: ±2.32 pp standardError 0.023232093333444247 ### Source [Original source](https://epoch.ai/data/benchmark_data.zip#file=gpqa_diamond.csv) Source epoch ai Original model label WizardLM-2-8x22B Source version GPQA diamond Evaluated 2025-01-27T00:00:00.000Z Observed 2026-10-22T03:00:05.281Z Evidence SHA-256 `9ef0b938bdd8406b029d595983ee8ad3a80bd9d64b82605b91b5ce39ba650c39` Source locator `/rows/307` |  |  |  |  |  |
| [MAGI-Hard](https://benchize.com/benchmarks/eq-bench%3Amagi-hard.md) | 59.16% | — | 9 / 14 | [eq bench · magi-hard](https://eqbench.com/eqbench-v2.js) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 59.16 \| percent \| overall \| Higher is better \| ### Configuration interactiveAdjustments none protocol magi-hard sourceModelVersion microsoft/WizardLM-2-8x22B ### Source [Original source](https://eqbench.com/eqbench-v2.js) Source eq bench Original model label microsoft/WizardLM-2-8x22B Source version magi-hard Observed 2026-10-22T05:10:05.138Z Evidence SHA-256 `088fe07c18b4347c82113934c58bb801cdb21adfe80852c7d0854e9818749dcd` Source locator `/leaderboardDataMagi/115` |  |  |  |  |  |
| [MATH Level 5](https://benchize.com/benchmarks/epoch-ai%3Amath_level_5.md) | 25.74% | SE ±0.87 pp | 20 / 21 | [epoch ai · MATH level 5](https://epoch.ai/data/benchmark_data.zip#file=math_level_5.csv) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 0.25736404833836857 \| score_fraction \| Best score (across scorers) \| Higher is better \| \| `mean_score` \| 0.25736404833836857 \| score_fraction \| mean_score \| Higher is better \| ### Configuration benchmarkVersion MATH level 5 reasoningBudget Not reported Reasoning effort Not reported sourceModelVersion WizardLM-2-8x22B ### Reported statistics Uncertainty Standard error: ±0.87 pp standardError 0.008719521575190186 ### Source [Original source](https://epoch.ai/data/benchmark_data.zip#file=math_level_5.csv) Source epoch ai Original model label WizardLM-2-8x22B Source version MATH level 5 Evaluated 2025-01-27T00:00:00.000Z Observed 2026-10-22T03:00:05.281Z Evidence SHA-256 `041b4bcb991a81d66152327113ef6d38e64cd8cce93d1e6ec821d1b9e47c1de3` Source locator `/rows/90` |  |  |  |  |  |
| [EQ-Bench v2](https://benchize.com/benchmarks/eq-bench%3Aeqbench-v2.md) | 77.91% | — | 8 / 18 | [eq bench · eqbench-v2](https://eqbench.com/eqbench-v2.js) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 77.91 \| percent \| overall \| Higher is better \| ### Configuration interactiveAdjustments none protocol eqbench-v2 sourceModelVersion microsoft/WizardLM-2-8x22B ### Source [Original source](https://eqbench.com/eqbench-v2.js) Source eq bench Original model label microsoft/WizardLM-2-8x22B Source version eqbench-v2 Observed 2026-10-22T05:10:05.138Z Evidence SHA-256 `088fe07c18b4347c82113934c58bb801cdb21adfe80852c7d0854e9818749dcd` Source locator `/leaderboardDataEqbench/123` |  |  |  |  |  |
| [NC Bench](https://benchize.com/benchmarks/nc-bench%3Aoverall.md) | 71.45% | — | 140 / 144 | [nc bench](https://www.nc-bench.com/models) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 71.45 \| percent \| reported \| Higher is better \| \| `tooling` \| 91.22 \| percent \| category \| Higher is better \| \| `creative_writing` \| 79.06 \| percent \| category \| Higher is better \| \| `language` \| 78.05 \| percent \| category \| Higher is better \| \| `utility` \| 67.14 \| percent \| category \| Higher is better \| \| `reasoning` \| 60.41 \| percent \| category \| Higher is better \| \| `text_editing` \| 88.13 \| percent \| category \| Higher is better \| \| `rule_following` \| 28.27 \| percent \| category \| Higher is better \| \| `hallucination` \| 79.28 \| percent \| category \| Higher is better \| ### Configuration Reasoning effort unknown Model configuration WizardLM 2 8x22b thinkingEnabled Not reported ### Source [Original source](https://www.nc-bench.com/models) Source nc bench Original model label WizardLM 2 8x22b Observed 2026-10-22T06:10:03.426Z Evidence SHA-256 `ec261022458543624b08a136d01f5a2f0666cd52ffb49c64d6ca983e8543d1b5` Source locator `/rows/190` |  |  |  |  |  |

---
Cite as: WizardLM-2 8x22B — Benchize, https://benchize.com/models/wizardlm-2-8x22b (retrieved 2026-10-22)