LLM AtlasLLM AtlasSearch models

DeepSeek

DeepSeek-V3.2-Exp

Overall 71Open weightopen-weight

DeepSeek's MoE flagship models (671B total, 37B activated params) with 128K context, strong on math, code, and reasoning benchmarks.

Last verified: 2026-03-29Confidence: HighSources: 4
textreasoningcodeopen-source
Input price
Not applicable
Output price
Not applicable
Context window
128,000
Max output
16,384
Release date
2024-12-26
Access
open-weight, self-hosted, hosted
License
MIT
Last verified
2026-03-29

Capability profile

Radar view of the model's practical strengths. This chart is backed by textual summaries below for crawlability.

Benchmark summary

This entry has source-backed metadata, but benchmark coverage is still sparse, so the page emphasizes published specs and deployment fit.

No benchmark series is attached to this model yet. Source links and product metadata are available below.

Strengths

  • • Source-backed listing for generalist reasoning buyers
  • • Useful when deeper benchmark profiling is still incomplete

Trade-offs

  • • Detailed benchmark coverage is still limited for this entry
  • • Some pricing or capability fields may remain unpublished

Crawlable benchmark analysis

DeepSeek-V3.2-Exp is positioned as a generalist reasoning model with published scores that emphasize its practical fit for buyers evaluating the entry.

Published scores highlight reasoning 70/100, coding 65/100, enterprise readiness 70/100, vision 55/100, speed 80/100, and safety 72/100.

Pricing is not applicable for this self-hosted or open-weight entry. With a context window of 128,000 tokens, it supports large-document analysis and retrieval workflows.

Benchmark coverage is still limited for this entry, so this section focuses on published metadata and deployment fit.

Sources

Provider and distribution links used to verify this model record.

Last verified: 2026-03-29

Related models

OpenAI

GPT-5.4

OpenAI

OpenAI's GPT-5.4, the most capable and efficient frontier model for professional work. First general-purpose model with native computer-use capabilities. Combines industry-leading coding from GPT-5.3-Codex with improved agentic workflows.

Score 933 sources
textreasoningtool-usevisionapihosted
Context
1,000,000
Input
$0.005/1K tok
Output
$0.02/1K tok
Coverage
Full profile
View analysis

Anthropic

Claude Sonnet 4.6

Claude 4.6

Anthropic's current Sonnet tier for fast frontier reasoning, coding, and long-context agent work.

Score 923 sources
textvisionreasoningcodetool-useapihosted
Context
1,000,000
Input
$0.003/1K tok
Output
$0.02/1K tok
Coverage
Full profile
View analysis

Anthropic

Claude Opus 4.6

Claude 1M

Anthropic's most intelligent Claude model for complex agents, coding, and deep reasoning, with 1M token context and 128K output.

Score 913 sources
textvisionreasoningapihosted
Context
1,000,000
Input
$0.005/1K tok
Output
$0.03/1K tok
Coverage
Full profile
View analysis