DeepSeek V3
DeepSeek · released Dec 25, 2024 · deepseek-ai/DeepSeek-V3
- Type
- Open weights
- Params
- 685B
- Context
- 164K
37B active per word · about 123K words of context
Our take
Written Sep 1, 2026DeepSeek V3 is a large text-only model with a mixture-of-experts design that keeps only 37 billion of its 685 billion parameters active for each token. Released in late 2024, it scores well on coding tasks and offers a 163,840-token request limit, though its licence terms remain unverified.
Choose this for budget coding workloads where its measured programming scores meet your bar, or for long-context text tasks up to 163,840 tokens. It suits cost-sensitive inference through competitively priced routes. Skip it if you need image, video or audio handling, if unverified licence terms pose legal risk, or if maths-heavy work dominates — its maths score sits well below its own coding and general-chat baselines.
The case for it
- Coding performance runs 29 points above its own general-chat baseline on the arena leaderboard we track, with a 49.6% pass rate on contamination-free competitive programming.
- Only 37 billion parameters active per token from 685 billion total — roughly one in nineteen engaged at once.
- Input pricing on some routes undercuts other tracked offers for this model.
The case against it
- Maths is a clear gap within its own profile, scoring 48 points below its general-chat baseline and 77 below its coding score.
- Text-to-text only; no image, video or audio input or output.
- Licence field undisclosed in our data — commercial terms unverified despite open weights, and output pricing varies sharply by provider.
How good is it?
An open text model for everyday questions and writing, though it trails most models on coding tasks.
- writing and completing codeArena Coding · 127th of 168
EverydayGeneral questions and everyday reasoning
Arena Text (overall)120th of 168 · 1358
CodingWriting and fixing code on its own
Arena Coding127th of 168 · 1387
AgenticPlanning, calling tools, staying on task
Not yet scored on Arena Agent.
WritingDrafting and rewriting prose
Arena Creative Writing102nd of 168 · 1349
Arena Creative Writing is the only board that has scored it for this.
These tests check whether a model follows instructions — a precondition for all the work above, but not a measure of how well that work is done.
Every published score for this model7 scoresEvery figure we hold, from 7 boards, with who ran it and a link to the source — including the boards no rating above is built on.
Can you run it yourself?
GeForce RTX 4090 · 24 GB
Too large for this card. The weights do not fit even with part of them offloaded to system memory.
Apple M1 Pro (16-core GPU) · 32 GB
Too large for this card. The weights do not fit even with part of them offloaded to system memory.
Radeon RX 7900 XT · 20 GB
Too large for this card. The weights do not fit even with part of them offloaded to system memory.
Memory use by level
Against a 24 GB card.
What is quantisation? →This model on every device we track71 devicesThe Q4 build most people download, on each device: what the weights come to, how much context the memory leaves, and whether it runs. Smallest device that runs it first.
Check against your own machine → · Where to rent it hosted →
Or rent it from someone else
Prices checked 2 hours ago — each listing carries its own date.
The only listing at 164K of context — the other 3 in the table below are not like-for-like. 2 cheaper rows there are outside that comparison: a different quantisation or a different context length.
- per 1M tokens
- $0.26 in / $1.03 out
- Context served
- 164K
- Throughput
- Not measured
| Provider | In / out per 1M tokens | Context | Throughput | Trains on prompts | Logs prompts | Zero retention |
|---|---|---|---|---|---|---|
| DeepInfrafp4Direct and through OpenRouter | $0.32 / $0.89checked 2 hours ago | 164K16K max reply through OpenRouter | 12 tok/sthrough OpenRouter | DirectUnknownThrough OpenRouterNo | DirectUnknownThrough OpenRouterNo | DirectUnknownThrough OpenRouterConfirmed |
| Novita AIDirect | $0.89 / $0.89checked 2 hours ago | 64K | not measured | Unknown | Unknown | Unknown |
| OpenRouterOpenRouter's own listing | $0.26 / $1.03checked 2 hours ago | 164K | not measured | Unknown | Unknown | Unknown |
| StreamLakeThrough OpenRouter | $0.26 / $1.03checked 2 hours ago | 128K16K max reply | 31 tok/s | No | Yesunknown period | Unknown |
DeepInfra
Novita AIAcross the 4 listings we hold: 2 say they do not train on prompts (1 of them only through OpenRouter), 0 say they do and 2 do not say. 1 appears in the zero-retention registry we check (only through OpenRouter); the rest are unknown to us.
What each host's API supports
From the parameter list each endpoint publishes. Streaming is omitted: nothing we hold reports it, for any model.
| Provider | Tool calling | JSON output | Strict schema |
|---|---|---|---|
| DeepInfrafp4Direct and through OpenRouter | ✓ | ✓ | ✓ |
| Novita AIDirect | |||
| OpenRouterOpenRouter's own listing | ✓ | ✓ | ✓ |
| StreamLakeThrough OpenRouter | ✓ | ✓ | ✓ |
Tool calling: 3 of 4 listings say yes, 1 publishes no parameter list. JSON output: 3 of 4 listings say yes, 1 publishes no parameter list. Strict schema: 3 of 4 listings say yes, 1 publishes no parameter list.
Models people weigh against DeepSeek V3
When we formed this view
Recent changes
What moved
first indexed by our pipelineEach date is the day we first saw the change, or the day the maker announced it.
What we do not know about this model yet
- 1 of 4 listings publishes no parameter list, so what its API accepts is unknown to us.
- Nothing we hold says whether an endpoint streams, so we do not show it either way.
- 2 of 4 listings do not say whether they train on prompts, and 1 answers only through OpenRouter, not for its own listing.
- We hold no cached-input rate for any of its listings.
- We hold no batch or off-peak rate for any of its listings.
- We hold a decode speed for it, but no prompt-processing (prefill) figure, so how long the input side of a job takes is unknown to us.
Licence and identifiers
What the licence allowsWe hold no licence record for this model. Inside are the identifiers you need to pull it — its Hugging Face repo where we have one, our slug and a machine-readable card.
Licence
We hold no licence record for this model yet.
Identifiers
- Hugging Face
- deepseek-ai/DeepSeek-V3
- Architecture
- Mixture of experts
- Takes in, gives back
- Text in, text out
- Catalogue slug
- deepseek-deepseek-v3