See how ChatGPT, Perplexity and Google AI Overviews describe you today
Free AI Visibility AuditDeepSeek V3.1 Terminus is a text-only large language model from DeepSeek, released as a maintenance update to DeepSeek V3.1 that keeps the original model's capabilities while fixing reported problems with language consistency and agent behaviour.
DeepSeekVerified 24/09/2026Released 22/09/202501 / decision
Each figure sits next to the middle of the field, so it can be read as dear or cheap, wide or narrow, rather than floating on its own. Compared against the 126 models we track, not against an absolute standard.
Capability
Not measured
No published benchmark scores yet
Input / 1M tokens
$0.27
field median $0.30
Output / 1M tokens
$1.00
field median $1.25
Context
164K
field median 524K
02 / overview
Four questions, answered separately, because somebody arrives at one of them rather than at the top of the page.
The job it was built for, and the job it is not for.
DeepSeek V3.1 Terminus is a general-purpose text model, built as a corrective release rather than a new generation. The work went into two specific complaints from users of DeepSeek V3.1: output drifting between languages, and unreliable behaviour when the model is driving tools in an agent loop. It handles text only, so there is no image, audio or video input.
Where it sits in its line, and when a sibling is the better pick.
DeepSeek V3.1 Terminus first appeared in September 2025, sitting directly alongside DeepSeek V3.1 as its refreshed version rather than a separate line. It is the right pick over plain V3.1 in almost any case, particularly if you hit mixed-language output or flaky tool calls. It is the wrong pick if you need a model that reads images or audio, or if you have pinned and validated a pipeline against the earlier V3.1 build and cannot revalidate.
The API, the apps it powers, and what its limits let you do.
DeepSeek V3.1 Terminus is reached through the DeepSeek API and through aggregators that route to it, and it also sits behind DeepSeek's own chat app. The context window is large enough to hold long documents, extended transcripts or a full agent run's history in a single call, and the output ceiling is generous enough for long-form drafts and sizeable code files. Everything in and out is text.
What changes because this exists, or why it does not.
Agent work is where small inconsistencies compound, and a model that occasionally answers in the wrong language or mishandles a tool call is hard to build on. DeepSeek V3.1 Terminus makes the existing V3.1 behaviour steadier without asking teams to move to a different model family or rewrite prompts. It is an unremarkable release by design, and that is the point: the same model, fewer surprises.
Follows DeepSeek V3.1. Superseded by DeepSeek V3.2 Exp. See the whole line.
03 / evidence
Split by category, so a strong number never hides a thin evidence base. Verified means we read it on the benchmark's own published results; a provider's claim about its own model is shown and labelled rather than dropped.
No published benchmark scores for this model yet.
We publish a score only where we can link the result to where it was published. Until a benchmark result for this model exists in a source we read, this section stays empty rather than being filled with a provider's marketing figure.
How we decide what counts as evidence04 / ledger
Every published row, grouped by category, each compared with the best published score on the same benchmark. 'Is 64% good' is a question nobody can answer; '26 points behind the leader' is one anybody can.
Nothing in the ledger yet.
Each row here carries a score, the benchmark it came from, what the leading model scored on the same test, and a link to the published result. Rows appear as results are published and read.
How we decide what counts as evidence05 / capability
Where this model is strong, and against how many peers. Ranks are against models with evidence in that category, not against everything we track: ranking against models nobody tested would rank who published, not who is better.
No category scores to shape yet.
A category score is the weighted mean of the benchmarks published for it. With no published rows there is nothing to average, and an empty chart drawn at zero would say something false.
How we decide what counts as evidence06 / cost
List API rates as last read from the provider, with the source on every row, plus every change we have recorded since we started tracking it.
| Charge | Price | Unit | Read on | Source |
|---|---|---|---|---|
| Input | $0.27 | per 1M tokens | 2026-09-24 | Check |
| Output | $1.00 | per 1M tokens | 2026-09-24 | Check |
DeepSeek V3.1 Terminus is one of the cheaper capable models to run, with output charged at a few times the input rate, which is the usual shape. Against the frontier models from the larger US labs it sits well below them on cost, which is why it shows up in high-volume assistant and agent workloads where per-call spend matters more than the last few points of quality.
| Workload | Input / month | Output / month | Cost |
|---|---|---|---|
| A small product team | 20M tokens | 5M tokens | $10.40 |
| A busy support assistant | 200M tokens | 40M tokens | $94.00 |
| A document pipeline | 1000M tokens | 100M tokens | $370.00 |
List API rates, no caching and no batch discount, which both providers offer and which change the answer a great deal. Treat these as the ceiling, not the bill.
07 / specs
As listed by the provider's own catalogue and re-read every few hours. Anything absent is absent there too.
| Context window | 163,840 tokens |
|---|---|
| Maximum output | 32,768 tokens |
| Modalities | text |
| Released | 22/09/2025 |
| Status | Current |
| Catalogue identifier | deepseek/deepseek-v3.1-terminus |
08 / lineage
What this model replaced, what replaced it, and what else its provider has in the field.
Also from DeepSeek
09 / line
Every model in this family in release order, so a page from eight months ago says in one glance that two newer ones exist.
Came before
← DeepSeek V3.1Came after
DeepSeek V3.2 Exp →Ordered by release date and worked out from the naming, so a new member slots in as soon as its page exists. A retired model keeps its page and its place in the line.
10 / notes
What this model changes for a brand trying to be cited in AI answers, and every change we have logged since it launched.
What it changes for you
DeepSeek models are cheap enough to sit underneath high-volume answer surfaces, so a share of the AI answers describing your category will be generated by this one rather than by a frontier model. That means your brand needs to be legible in plain text, with clear product descriptions, pricing pages and comparison content that a text-only model can read without help from images or diagrams. If your positioning only exists in a screenshot or a product video, DeepSeek V3.1 Terminus cannot cite it.
Most buyers meet DeepSeek V3.1 Terminus through DeepSeek's consumer chat app, where questions about tools and vendors get answered directly. Developers meet it through the DeepSeek API or through a routing layer that exposes it alongside other models. It also turns up inside third-party assistants and search products that pick DeepSeek for its low running cost.
Nothing published here yet. Changes appear within hours of a provider announcing them.
11 / questions
The things people ask about this model, answered from what is on this page rather than from anywhere else.
Models change what gets cited. We measure whether AI answers name your brand or your competitors across every assistant, and show you what to change.