DeepSeek V4 Pro
Open weightsReasoningReasoningTool use
1Mcontext
$0.66 / $1.98per 1M tokens (in / out)
1.6Tparameters
122stories in our feed
Specs
hover a value to see its sourceContext window1M
Max output384K
Parameters1.6T
Knowledge cutoff—
Licencemit
Commercial useCommercial OK
InputText
OutputText
HF downloads (30d)335.6K
Familydeepseek-thinking
API price
USD per 1M tokensInput$0.66
Output$1.98
Cached input$0.022
Change history
2026-10-20 Context 1M → 1Mmodels.dev
2026-10-03Price $0.435/$0.87 → $0.66/$1.98models.dev
2026-10-09
StoreBench Live-Commerce Agent EvaluationarXiv: DeepSeek
2026-10-02
AI芯片2027年可支持数亿智能体Andrew Curran
2026-10-01
LLM Agents Evade Oversight in Multi-Agent SystemsarXiv: DeepSeek
2026-09-29
Analog Design Bench 发布:50 个模拟电路任务测试智能体长时程设计能力arXiv: DeepSeek
2026-09-24
2026-09-09
中国开源模型满足数据本地化需求OpenRouter
2026-09-02
Harness-of-Harness Framework for Autonomous CodingarXiv: DeepSeek
2026-08-28
Coding Agents Trade-offs AnalysisDeepLearning.AI
2026-08-22
2026-08-21
2026-08-15
2026-08-13
2026-08-12
2026-08-07
递归合成框架RST:低成本生成长时程终端任务arXiv: DeepSeek
2026-08-03
2026-07-31
DeepSeek-V4-Flash-0731仅升级API,V4-Pro即将发布深度求索 DeepSeek
GPT 5.6 Luna 降价,AI 应用公司迎来成本拐点orange.ai
2026-07-30
CodeSpec:面向智能体长周期特性开发的双可执行规范方法arXiv: DeepSeek
2026-07-24
2026-07-22
AI实验室是否刻意优化“骑自行车的鹈鹕”?系统测试未发现证据Simon Willison’s Weblog
2026-07-19
阿里云 Qwen 3.8 max preview 开源权重模型预览上线,2.4T参数Paul Couvert
2026-07-14
高盛分析LLM商品化:中国模型定价远低于美国,价格战加剧AI Will
2026-07-13
2026-07-08
2026-07-06