Skip to content
ModelCap

DeepSeek V3

All-versions archive #112Arena #181License unverified

DeepSeek·deepseek/deepseek-chat

DeepSeek V3 is a language model published by DeepSeek. 3 OpenRouter endpoints are published in the latest snapshot. It accepts text input and returns text. The published context limit is 164K tokens. The catalogue declares support for tool use, structured outputs.

ModelCap Index · All-versions archive
52.6#112
Measured · 82% support · score interval 52.6–59.4
Market Gravity
42.1
Secondary market signal

Superseded by DeepSeek V3.2. It no longer holds a current-board position; its all-versions archive rank is #112.

Input
$0.257
per 1M tokens
Output
$1.03
per 1M tokens
Context
164K
tokens
Providers
3
endpoints
Gateway spend
latest · Vercel share

Providers

Uptime measured over the last 30 minutes
ProviderInputOutputContextQuantUptime
DeepInfrafp4Cheapest$0.32$0.89164Kfp497.02%
StreamLake$0.257$1.03128K99.32%
Novitafp8$0.40$1.3064Kfp899.93%

Hugging Face Inference Providers

Exact repository identity

Provider economics and telemetry published by the Hugging Face Router. These are operational signals, not benchmark results.

ProviderInputOutputContextTTFTThroughputFeatures
deepinfra
live
$0.32$0.89164K578 ms27.9 tok/stools · structured
novita
live
$0.40$1.3064K1160 ms33.2 tok/stools

Reported evaluation leads (11)

Excluded from score

Structured rows reported in this exact model repository at the pinned revision below. They are useful research leads, but remain quarantined until dataset revision, harness, configuration and identity pass ModelCap’s independent admission review.

Dataset / taskMetricValueRevisionProvenance
Idavidrein/gpqa
diamond
58.2071
Unpinned
not published
Reported
Quarantined
LEXam-Benchmark/LEXam
open_question
52.53
Unpinned
not published
Reported
Quarantined
LEXam-Benchmark/LEXam
mcq_4_choices
46.57
Unpinned
not published
Reported
Quarantined
openai/gsm8k
gsm8k
89.3
Unpinned
not published
Reported
Quarantined
thamilvendhan/signalbench
src
0.5667
Unpinned
not published
Reported
Quarantined
thamilvendhan/signalbench
time
0.8333
Unpinned
not published
Reported
Quarantined
thamilvendhan/signalbench
access_deny
0.5833
Unpinned
not published
Reported
Quarantined
thamilvendhan/signalbench
memory_label
0.5833
Unpinned
not published
Reported
Quarantined
thamilvendhan/signalbench
injection
0.5833
Unpinned
not published
Reported
Quarantined
thamilvendhan/signalbench
bot_policy
0.25
Unpinned
not published
Reported
Quarantined
TIGER-Lab/MMLU-Pro
mmlu_pro
64.4
Unpinned
not published
Reported
Quarantined

Other versions (4)

ModelCap Index · All-versions archive

Methodology
52.6
Measured · 2 public benchmark observations across 2 boards
All-versions archive
#112
Public rank basis
Measured (measured)
Index support
82.4% · strong
Index interval
52.6–59.4
Rank posterior
Not available
Top 5 / Top 10 probability
Not available / Not available
Posterior as of
Snapshot timestamp unavailable
Identity binding
Exact catalogue product · deepseek/deepseek/deepseek-chat/fp8 · aggregates disclosed benchmark configurations · endpoint configuration not claimed · artifact metadata deepseek-ai/DeepSeek-V3@e815299b0bcbac849fa540c768ef21845365c9eb (not the evaluation revision)
Direct benchmark observations
2
Breadth-adjusted BES point estimate
56.0
Bayesian Evidence Score · measured evidence input only

BES normalizes admitted public benchmark evidence for measured Index rows. Its legacy score and observables prior do not define the public language rank.

Observed capability
56.0
Evidence status
confirmed · 83% mass
General preference
56.7
Coding
53.3
Agents & tools
Reasoning
Evidence breadth
92%
Evidence coverage
82%
Market and catalogue signals · secondary, never the language Index rank
Usage (OpenRouter popularity)
21.0
Liquidity (providers × uptime)
89.9
Open reach (HF downloads)
Surface (context / tools / modalities)
46.8
Freshness
10.6
  • · OpenRouter popularity #90
  • · 3 live providers
  • · Catalogue: context 163840, tools
  • · First seen 2024-12-26

These adoption and deployment observations remain context only. They do not change this model's ModelCap Index score or rank.

BES evidence selected · measured input only

The bounded 0–100 capability score blends each source’s competitive placement with its published achievement, then combines capability families. Evidence breadth adds a modest uncertainty adjustment; price and popularity are not part of benchmark-led rank.

Deployment readiness

Metadata index
72.1
partial · 82% metadata coverage
Not capability
Artifact reproducibility
85.0
Access & legal clarity
32.5
Deployability
99.8
Evaluation provenance
51.0

Missing: artifactReproducibility.base-lineage, accessLegalClarity.declared-license, accessLegalClarity.resolvable-license-terms, evaluationProvenance.evaluation-dataset-revisions.

A metadata completeness and deployability index—not a safety certification, quality grade, or production-readiness claim.

Market activity

Venue coverage
Market Gravity
42.1
OpenRouter weekly popularity
#90 of 298

Market Gravity uses OpenRouter’s full-catalogue popularity order. Sparse Vercel top-ten observations appear only when published and do not affect the score.

Arena preference

Status
Ranked
Source
Arena
Official rank
#181 of 385
Rating
1358.6
95% confidence interval
1354–1363
Votes
22K
Source snapshot
6 Aug 2026, 13:00 UTC
Category
overall
Identity match confidence
100%

Market Gravity breakdown

Usage55%
21.0

OpenRouter weekly popularity across the full model catalogue.

Liquidity25%
89.9

Independent providers versus the model's open or closed cohort, weighted by uptime.

Open reach15%
50.0

Hugging Face 30-day downloads for open models; neutral for closed models.

Freshness5%
10.6

Time since first public availability, on a six-month half-life.

Specification

Context window
164K tokens
Max output
16K tokens
Inputs
Text
Outputs
Text
Tokenizer
DeepSeek
Cached input tokens
Not offered
First seen on OpenRouter
26 Dec 2024
Knowledge cutoff
2024-07-31

Capabilities

  • Not supported: Reasoning
  • Supported: Tool use
  • Supported: Structured output
  • Supported: Response format
  • Not supported: Moderated

Weights & access

License unverified. A weight repository exists, but ModelCap could not verify its license classification from the current source metadata.

Access
License unverified
Pinned revision
e815299b0bcb
Parameters
684.5B
Architecture
DeepseekV3ForCausalLM
Model type
deepseek_v3
Hugging Face downloads (30d)
1.2M
Downloads (all time)
19.3M
Likes
4K
Repository updated
27 Mar 2025, 04:01 UTC