LIVE
Introducing Cosmos 3 Edge20/07/26 · Hugging Face|Kimi K3, Qwen 3.8, and Anthropic's (Potential) Unravelling20/07/26 · Anthropic|At SIGGRAPH, NVIDIA Advances Graphics and Simulation With Agentic and Physical AI20/07/26 · NVIDIA|China's open-weights AI strategy is winning20/07/26|Bristol Myers Squibb Building Life Science Industry’s Most Advanced AI Factory on NVIDIA Vera Rubin20/07/26 · NVIDIA|Safety and alignment in an era of long-horizon models20/07/26 · OpenAI|Claude Fable produced a counterexample to the Jacobian Conjecture20/07/26 · Anthropic|Apply for Anthropic’s AI for Science rare disease research grants20/07/26 · Anthropic|AI advice made people less accurate but more confident – sudy19/07/26|Vincentwei1021/video-shotcraft: AI video skill for Claude Code & Codex — cinematic product videos with Remotion: 106 shot recipe cards, 161 motion previews, a production-ready template19/07/26 · Anthropic|Claude Code uses Bun written in Rust now19/07/26 · Anthropic|OpenAI reduces Codex Model Context Size from 372k to 272k19/07/26 · OpenAI|Introducing Cosmos 3 Edge20/07/26 · Hugging Face|Kimi K3, Qwen 3.8, and Anthropic's (Potential) Unravelling20/07/26 · Anthropic|At SIGGRAPH, NVIDIA Advances Graphics and Simulation With Agentic and Physical AI20/07/26 · NVIDIA|China's open-weights AI strategy is winning20/07/26|Bristol Myers Squibb Building Life Science Industry’s Most Advanced AI Factory on NVIDIA Vera Rubin20/07/26 · NVIDIA|Safety and alignment in an era of long-horizon models20/07/26 · OpenAI|Claude Fable produced a counterexample to the Jacobian Conjecture20/07/26 · Anthropic|Apply for Anthropic’s AI for Science rare disease research grants20/07/26 · Anthropic|AI advice made people less accurate but more confident – sudy19/07/26|Vincentwei1021/video-shotcraft: AI video skill for Claude Code & Codex — cinematic product videos with Remotion: 106 shot recipe cards, 161 motion previews, a production-ready template19/07/26 · Anthropic|Claude Code uses Bun written in Rust now19/07/26 · Anthropic|OpenAI reduces Codex Model Context Size from 372k to 272k19/07/26 · OpenAI|
🇨🇳 China🧠 GeneralistReleased November 2025

DeepSeek V3.2

DeepSeek
Open-sourceFreeSelf-host
Verdict

A solid pick for self-hosting and full data control.

Overview

DeepSeek V3.2 introduces DeepSeek Sparse Attention (DSA) to cut training and inference cost on long contexts without meaningful quality loss. MIT-licensed, 685B parameters in a MoE design, it delivers solid reasoning (82.4% GPQA) at one of the lowest prices on the market. Ideal for volume and for teams that want to self-host and keep control of their data.

Skill profile

FrenchReasoningSpeedCreativitySafety
Verified benchmarks
GPQA82.4%
SWE-Bench67.8%

What do these scores mean?

GPQA82.4%excellent

PhD-level science questions (physics, chemistry, biology), with no tool access.

SWE-Bench67.8%very good

Resolving real GitHub issues under real conditions (SWE-bench Verified).

Strengths

  • Excellent at code
  • Has a free tier
  • Open-source and self-hostable
  • Competitive input pricing

Limitations

  • No native GDPR guarantee
  • Light safety filters

Who is it for

A good fit if…
  • you build with a coding agent
  • you want to control cost or self-host
Skip it if…
  • you handle sensitive EU data
  • you want strict guardrails

Ideal use cases

  • High-volume usage at low cost
  • Self-hosting and data control
  • Reasoning and code on a budget

Access & availability

Paid API per tokenFree tier availableSelf-hostable (open weights)

Key specifications

Context
131K
Input price
$0.28 $/M
Output price
$0.42 $/M
Speed
Price not auditedBenchmarks not audited

Estimate your monthly cost

Per-token API pricing
$11/ month with DeepSeek V3.2

For the same usage

Indicative estimate based on standard API rates (excluding caching, batch and volume discounts). Always check the official pricing before committing.

Privacy

Not GDPR-compliant
Advanced data · for experts
Architecture, modalities, detailed cost, full benchmarks
Architecture
MoE + DeepSeek Sparse Attention
Size
685B (MoE)
Cutoff date
Inputs
text
Outputs
text
License
mit
Hosting
Hallucination score
Estimated cost (API)
Typical exchange (~3k in / 1k out)
≈ $0.0013
1M in + 1M out
≈ $0.70
All benchmarks
Arena Elo
MMLU
GPQA82.4%
HumanEval
SWE-Bench67.8%
MATH

Similar models