MODEL PROFILE / EDITION 2026.09
DeepSeek V3.1
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...
No ordinal score is published because comparable evidence is incomplete. No rank delta is shown because there is no prior numeric snapshot.
Best for
Catalogue discovery; compare availability, context, and public token pricing before benchmark review.
Price & access
$0.55 input / $1.65 output per 1M tokens on OpenRouter.
Evidence summary
Availability and price come from OpenRouter; normalized frontend and coding evidence has not yet been reviewed.
Edition status
Tracked. Tracked in the September 2026 baseline.
Strengths
- Current model catalogue entry with a public API identifier and pricing metadata.
Tradeoffs
- No normalized frontend-quality score has been assigned.
- Popularity and recency are not treated as quality evidence.
Linked evidence
Evidence source 1 ↗Evidence source 2 ↗Use the same published data, filters, and side-by-side metrics.