DeepSeek V3.2 Exp
deepseek/deepseek-v3.2-exp- DeepSeek
- Family · DeepSeek V3
- Released 29 Sep 2025
- Open weights
- Text → Text
- Tools · JSON
Not on OSMEXO yet
Spec & list price
Context160K
Max output144K
Input / 1M$0.27
Output / 1M$0.411.5× input
Scores · standing
Intelligence16.6#165/489
Usage
Tokens / day1.5B▼ 22% 7d
Usage rank#172
Requests / day65.3K
Tool calls / req0.01
Per request · 30 days
Avg input7.0K0.5× median
Avg output672about median
Cache hit47%0.7× median
Reasoning share3.5%0.1× median
Tool-call failures0.2%0.1× median
Consumption
Peak 2.9B 7 OctLow 688.7M 31 OctRank today #172
News & trends
6 mentions this week ▼ 14%DeepSeek V3 family · videos, posts, articles · 30 days
Trending · 14 days · DeepSeek V3 family
- What Did DeepSeek V3's $5.6 Million Cost Leave Out? #deepseek #nvidia #ai #mixtureofexperts1.5K views
- Best provider for deepseek v3.2?17 upvotes
- Free Claude Code CLI Setup: Use DeepSeek V3 & Gemini API (1-Click Windows Installer)29 views
News · latest from DeepSeek
- Tell HN: Mistrap Pro includes DeepSeek v4.1 Flash 1017docs.mistral.ai · 1 Nov
- Show HN: DeepSeek v4.1 Flash-1017 Safety Evaluationtwitter.com · 26 Oct
- DeepSeek V4 Pro 0813 vs Claude Fable 5 on DeepSWE: Cost, Coding, and RoutingTogether AI · 17 Aug
- DeepSeek-V4 Flash 0731 vs GPT-5.6 Luna on DeepSWE: Cost and CodingTogether AI · 6 Aug
Benchmarks
bar = share of the top score · Benchmark data: Artificial AnalysisSpeed · median across providers, latest day
Output speed#321/33317 tok/s3.9× slower than median
First token#2642.22 s2.2× longer than median
#4 cheapest of 18 with a similar score
Value
Intelligence Index
Blended price per 1M tokens (log)
Family progress
Intelligence Index
- Family releases21.5
Technical details
- Tokenizer
- DeepSeek
- Knowledge cutoff
- 2025-07-31
- License
- MIT
- Parameters
- 685B
- Moderation
- none
- Downloads · likes
- 43.4K · 1.0K
Request parameters 11
- tools
- tool_choice
- response_format
- structured_outputs
- reasoning
- include_reasoning
- temperature
- top_p
- top_k
- max_tokens
+1 more
- frequency_penalty
When to use
Best for
- Cost-efficient long-context reasoning
- Tool-calling agents over long inputs
- Coding and math tasks
- Self-hosted general models (open weights)
Think twice for
- Production use that needs a stable release (experimental)
- Image understanding
- Newer projects (usage is declining)
Research papers
Same family
DeepSeek V3- 13.7
- 21.5
- 8.5
- 9.7
Alternatives
other vendors · similar capability- 13.1
- S19.5
- 20.6
- 15.6
Want DeepSeek V3.2 Exp on OSMEXO?
Sign in to request it. Requests are reviewed; a request does not guarantee availability.