All models

DeepSeek-V3.1

byDeepSeekDeepSeek· 21 Aug 2025
General purpose

To the surprise of everyone, DeepSeek launched another iteration in their V3 series. This model combines the reasoning of R1 with the non-reasoning component of V3, which is different from Qwen which is splitting its models again.

Specs
Params671B
LicenseMIT
Capability · Artificial Analysis
AA Index
21.0
Months Behind Frontier
8.5 mo
o1
05 Dec 2024
Adoption · Hugging Face
RAM score
Hugging Face Downloads
283.4K
last 30d
2.6M
all time
HF Likes
825

Relative Adoption Metric contextualizes downloads against the model's size bucket.

Inference · OpenRouter
Tokens/Day
below top 50 this week
Peak Tokens/Day
45.9B
05 Oct 2025
Peak Rank
#2
05 Oct 2025

OpenRouter publishes daily token totals for its ~50 most-served models. Days missing from the chart mean DeepSeek-V3.1 fell below that cutoff — not zero usage. OpenRouter logs usage separately per dated model version and per variant (like ":free"); the hub sums them into one daily total per model.

Related Models