Automated page. Written and translated automatically by AI and checked against the sources by software; no human editor. SI-IQ is an index built from public benchmark results, not an IQ test. How it works

How smart is DeepSeek-V3 (Mar 2025)?

85 SI-IQ · 90% interval 81–90
Not ranked: too little data

DeepSeek-V3 (Mar 2025) is a model from DeepSeek, released on Mar 24, 2025. Its SI-IQ is 85 (90% interval 81–90), estimated from 4 benchmark results by 2 independent evaluators.

This model has fewer than 5 independent benchmark results or fewer than 2 different evaluators, so it is shown but not ranked.

Facts

Developer: DeepSeek
Released: Mar 24, 2025
Access: Open weights (restricted licence)
Benchmarks used: 4 results from 2 evaluators

Ability by area (SI-IQ, 90% interval)

Every benchmark result we used

Each score is one value exactly as the source published it: for Epoch AI the best of its run settings, for OpenRouter the median hosting provider (the developer's own hosting when it ran the test). “Implied SI-IQ” is the ability that this single result points to; the overall SI-IQ is their precision-weighted average.

Benchmark	Area	Score	Implied SI-IQ	Evaluated by	As listed by the source	Retrieved	Source
MATH Level 5	Math	75.5%	83	Epoch AI	`DeepSeek-V3-0324`	Oct 18, 2026	Epoch AI Benchmarking Hub · data file
OTIS Mock AIME 2024–2025	Math	37.8%	88	Epoch AI	`DeepSeek-V3-0324`	Oct 18, 2026	Epoch AI Benchmarking Hub · data file
GPQA Diamond	Knowledge	67.6%	92	Epoch AI	`DeepSeek-V3-0324`	Oct 18, 2026	Epoch AI Benchmarking Hub · data file
GPQA Diamond (OpenRouter run)	Knowledge	64.6%	91	OpenRouter	`deepseek/deepseek-chat-v3-0324` Crusoe, DeepInfra, GMICloud, Novita, SiliconFlow · median over providers (5 of 5 rows)	Oct 18, 2026	OpenRouter benchmark datasets · data file
τ²-bench Airline (OpenRouter run)	Agents	43.4%	80	OpenRouter	`deepseek/deepseek-chat-v3-0324` DeepInfra · median over providers (2 of 2 rows), lower median	Oct 18, 2026	OpenRouter benchmark datasets · data file

Human preference (Arena)

On the Arena leaderboard, where people vote between two anonymous answers, DeepSeek-V3 (Mar 2025) has a rating of 1,395 (1,391–1,399) from 44,973 votes (published Oct 15, 2026). Human preference is a different measure from benchmark ability and is not part of SI-IQ. Arena (Hugging Face)

Computed with SI-IQ method version 1.0.0. Read the full methodology

Data sources: Arena leaderboard dataset on Hugging Face (CC BY 4.0); Epoch AI, ‘Capabilities & benchmarking’ (CC BY 4.0); Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter..