DeepSeek · mit

DeepSeek-V3.2-Speciale

DeepSeek-V3.2-Speciale by DeepSeek appears in 1 source with Reasoning at 43.22. Best read for Code quality, Math, Open weights.

CreatorDeepSeek

Release date2025-12-01

Knowledge cutoffNot published

ContextUnknown

Input priceUnknown

Output priceUnknown

Modalitytext

CountryCN

Metrics

All source-backed metrics

Open in leaderboard

LLM Stats Rank62ranking · source_rank

Reasoning43.22reasoning · index_reasoning

Math45.83math · index_math

Coding22.64coding · index_code

Vision24.53multimodal · index_vision

Tool calling12.62tool_calling · index_tool_calling

AIME 202596 %math · aime_2025_score

SWE-bench Verified73.1 %coding · swe_bench_verified_score

Code Arena832.32 %coding · coding_arena_score

Humanity Last Exam30.6 %reasoning · hle_score

Toolathlon35.2 %tool_calling · toolathlon_score

Parameters685,000,000,000 paramsmodel · params

Evidence

Citations and source overlap

LLM StatsPublic full LLM leaderboard advanced explorer snapshot, including ranking, benchmark indexes, pricing, speed, context, license, and model metadata.Retrieved 2026-06-01

FAQ

How should I read this profile?

Treat this as a source-backed model dossier, not an EvalKit-run verification. The public values are replicated from linked sources and kept source-scoped.

Is DeepSeek-V3.2-Speciale verified by EvalKit?

No. EvalKit currently shows 0 verified rows until real run evidence exists.

Why can metrics disagree?

Different sources test different tasks, dates, prompts, and aggregation methods. EvalKit keeps those differences visible instead of merging them into a fake universal score.