DeepSeek: DeepSeek V3.1
Released Aug 21, 2025Try it on:
DeepSeek-V3.1 is a large hybrid reasoning model (671B parameters, 37B active) that supports both thinking and non-thinking modes via prompt templates. It extends the DeepSeek-V3 base with a two-phase long-context...
Best overall rank
-
Across all subsets · overall
Total arena votes
0
Across 0 (subset, category) entries
Providers on OR
0
None yet
First seen
-
Awaiting first snapshot
Trend
No snapshot history yet - trends appear after the first daily refresh.
Pricing
Per token
- Input (per 1M)
- $0.210
- Output (per 1M)
- $0.790
- Cache Read (per 1M)
- $0.130
Capabilities
Modality
text->textContext
164K tokens
Tokenizer
DeepSeek
Knowledge cutoff
2025-03-31
Max output
33K tokens
Moderation
Unmoderated
Hugging Face
Supported parameters
frequency_penalty
include_reasoning
logit_bias
max_tokens
min_p
presence_penalty
reasoning
repetition_penalty
response_format
seed
stop
structured_outputs
temperature
tool_choice
tools
top_k
top_p
Providers
No provider endpoint data yet.
Per-category ranks
No leaderboard entries for this model.