# DeepSeek V3 > Source: Benchize (https://benchize.com) — https://benchize.com/models/deepseek-chat. When citing or reusing this content, credit "Benchize" and link to https://benchize.com/models/deepseek-chat. DeepSeek Canonical URL: https://benchize.com/models/deepseek-chat Last updated: Thu, 22 Oct 2026 05:30:58 GMT Reproduction and quotation must name Benchize, link to the canonical HTML URL, and retain individual source citations. DeepSeek ## Benchmarks | Benchmark | Score | ± | Rank | Source | | | --- | --- | --- | --- | --- | --- | | [GPQA Diamond](https://benchize.com/benchmarks/gpqa-diamond.md) | 71.21% | SE ±3.23 pp | 87 / 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` \| 0.7121212121212122 \| score_fraction \| Best score (across scorers) \| Higher is better \| \| `mean_score` \| 0.7121212121212122 \| score_fraction \| mean_score \| Higher is better \| ### Configuration benchmarkVersion GPQA diamond reasoningBudget Not reported Reasoning effort Not reported sourceModelVersion deepseek-chat ### Reported statistics Uncertainty Standard error: ±3.23 pp standardError 0.03225883512300997 ### Source [Original source](https://epoch.ai/data/benchmark_data.zip#file=gpqa_diamond.csv) Source epoch ai Original model label deepseek-chat Source version GPQA diamond Evaluated 2026-07-16T00:10:43.000Z Observed 2026-10-22T03:00:05.281Z Evidence SHA-256 `9ef0b938bdd8406b029d595983ee8ad3a80bd9d64b82605b91b5ce39ba650c39` Source locator `/rows/142` | | | | | | | [Creative Writing v2](https://benchize.com/benchmarks/eq-bench%3Acreative-writing-v2.md) | 81.19% | CI 80.76–81.61% | 6 / 18 | [eq bench · creative-writing-v2](https://eqbench.com/creative_writing_v2.js) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 81.19 \| percent \| overall \| Higher is better \| ### Configuration interactiveAdjustments none protocol creative-writing-v2 sourceModelVersion deepseek-ai/DeepSeek-V3 ### Reported statistics Uncertainty Confidence interval: 80.76–81.61% confidenceIntervalHigh 81.61 confidenceIntervalLow 80.76 ### Source [Original source](https://eqbench.com/creative_writing_v2.js) Source eq bench Original model label deepseek-ai/DeepSeek-V3 Source version creative-writing-v2 Observed 2026-10-22T05:10:05.638Z Evidence SHA-256 `264c33d486559c25267ca29861754404dadb674022dcdf6530388cd9951af31a` Source locator `/leaderboardDataCreativeWriting/11` | | | | | | | [OTIS Mock AIME 2024–2025](https://benchize.com/benchmarks/epoch-ai%3Aotis_mock_aime_2024_2025.md) | 48.89% | SE ±7.54 pp | 94 / 124 | [epoch ai · OTIS Mock AIME 2024-2025](https://epoch.ai/data/benchmark_data.zip#file=otis_mock_aime_2024_2025.csv) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 0.4888888888888889 \| score_fraction \| Best score (across scorers) \| Higher is better \| \| `mean_score` \| 0.4888888888888889 \| score_fraction \| mean_score \| Higher is better \| ### Configuration benchmarkVersion OTIS Mock AIME 2024-2025 reasoningBudget Not reported Reasoning effort Not reported sourceModelVersion deepseek-chat ### Reported statistics Uncertainty Standard error: ±7.54 pp standardError 0.0753592220347252 ### Source [Original source](https://epoch.ai/data/benchmark_data.zip#file=otis_mock_aime_2024_2025.csv) Source epoch ai Original model label deepseek-chat Source version OTIS Mock AIME 2024-2025 Evaluated 2026-07-16T00:10:43.000Z Observed 2026-10-22T03:00:05.281Z Evidence SHA-256 `b7c9da8c0587c82f1e0d814ceb496c24308380b997a16203b6e5d39eb07f4f50` Source locator `/rows/148` | | | | | | | [BuzzBench](https://benchize.com/benchmarks/eq-bench%3Abuzzbench.md) | 51.63% | CI 50.9–52.36% | 7 / 15 | [eq bench · buzzbench](https://eqbench.com/buzzbench.js) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 51.63 \| percent \| overall \| Higher is better \| ### Configuration interactiveAdjustments none protocol buzzbench sourceModelVersion deepseek-ai/DeepSeek-V3 ### Reported statistics Uncertainty Confidence interval: 50.9–52.36% confidenceIntervalHigh 52.36 confidenceIntervalLow 50.9 ### Source [Original source](https://eqbench.com/buzzbench.js) Source eq bench Original model label deepseek-ai/DeepSeek-V3 Source version buzzbench Observed 2026-10-22T05:10:06.154Z Evidence SHA-256 `4d0f19f1f87e08772b741a28b1ac39111587b832035dd1a783af3de67d51be1c` Source locator `/leaderboardDataBuzzbench/5` | | | | | | | [FinanceBenchmark v2](https://benchize.com/benchmarks/financebenchmark%3Av2.md) | 60.27% | CI 49.32–71.23% | 21 / 21 | [financebenchmark](https://finbenchmark.ai/) | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `pass_at_1` \| 0.60274 \| score_fraction \| reported \| Higher is better \| \| `accuracy` \| 0.593607 \| score_fraction \| reported \| Higher is better \| \| `knowledge` \| 1 \| score_fraction \| category \| Higher is better \| \| `analysis` \| 0.866667 \| score_fraction \| category \| Higher is better \| \| `quant` \| 0.266667 \| score_fraction \| category \| Higher is better \| \| `portfolio_mgmt` \| 0.5 \| score_fraction \| category \| Higher is better \| \| `credit_underwriting` \| 0.333333 \| score_fraction \| category \| Higher is better \| \| `banking_regulation` \| 0.4 \| score_fraction \| category \| Higher is better \| \| `derivatives_structuring` \| 0.666667 \| score_fraction \| category \| Higher is better \| \| `corporate_finance` \| 0.2 \| score_fraction \| category \| Higher is better \| ### Configuration harnessVersion 0.1.0 Reasoning effort unknown sourceRunId c4e5bb67-ca9f-4399-a3a4-2427c7f67150 taskSetVersion v2 ### Reported statistics Uncertainty Confidence interval: 49.32–71.23% confidenceInterval [0.4932,0.7123] confidenceMethod percentile bootstrap over tasks, n=5000, alpha=0.05 sampleSize 73 ### Source [Original source](https://finbenchmark.ai/) Source financebenchmark Original model label deepseek-chat Evaluated 2026-06-15T09:52:08.64329+00:00 Observed 2026-10-22T04:10:04.512Z Evidence SHA-256 `f9e7dab8846b092fc0befeca8c5cd02a8187c0f5b6b10c32ab2c7ab60f7af575` Source locator `/rows/25` | | | | | | --- Cite as: DeepSeek V3 — Benchize, https://benchize.com/models/deepseek-chat (retrieved 2026-10-22)