LLM Pricing
PricingLeaderboardsToolsProvidersReleasesGuides

© 2026 LLM Pricing

About
·Contact
·Privacy
·RSS
← Model list

DeepSeek-V3

deepseek·deepseek/deepseek-v3·Deprecated·Open weights·deepseek series
Open DeepSeek MoE chat model for coding, math, and general reasoning

DeepSeek-V3 is offered by 15 providers on this page. Public prices are shown for 12 of them. Its reference price is $1.25 per 1M tokens. The lowest paid channel is SiliconFlow at $0.25 / $1.00 per 1M, about 5.0× below the reference price. It also has 1 free ($0) channel; free tiers usually carry rate limits, and subscription-covered access bills $0 per token only after the subscription fee.

Artificial Analysis rates it 8.5 on the Intelligence Index, with 23 for coding and 0.8 for agentic tasks. Against its lowest blended price of $0.438 per 1M, that is roughly 19 index points per dollar, which is the value ratio the leaderboards rank on. Running one task of the Artificial Analysis suite costs about $0.0198, which reflects how many tokens its reasoning consumes rather than the unit price alone.

The context window is 131,072 tokens at the reference host, but hosts report different limits, from 64,000 to 164,000, so the usable window depends on the provider you pick. It supports tool use and structured output. The weights are open, so it can also be self-hosted or served through a gateway of your choice. Providers report the capability flags inconsistently, so verify a specific feature against the host you plan to use. Its training knowledge cuts off at 2024-07.

Specs & pricing

Input / output per 1M tokens
Reference price·Together AI
$1.25 / $1.25
Blended $1.25 · Cache read —
Lowest paid·SiliconFlowCloud
$0.25 / $1.00
Blended $0.44 · 5× spread
1 more $0 channels
Context
131,072
Output limit
131,072
Knowledge cutoff
2024-07
Released / updated
2024-12-26 / 2024-12-26
Capabilities
Reasoning✓ Tool use✓ Structured output✓ TemperatureAttachments
⚠ Providers report capability flags inconsistently
Modalities
Text
License
DeepSeek Model License
Weights
Hugging Face

Quality & performanceArtificial Analysis · Intelligence Index v4.3

Intelligence8.5
Value19
Coding23
Agentic0.8
Output speed—
TTFT—
Cost per task$0.02
Value formulaIQ 8.5 ÷ min blended $0.438 = 19

Quality is independently evaluated by Artificial Analysis. Speed/latency are model-level medians.

Available at 15 providers12 with public prices · 1 free

ProviderTierInputOutputCache readCache writeContextOutput limitStatus
iFlowGatewayFree——128,000 ⚠32,000
SiliconFlow
deepseek-ai/DeepSeek-V3
Cloud$0.25$1.00——164,000 ⚠164,000
SiliconFlow (China)
deepseek-ai/DeepSeek-V3
Cloud$0.25$1.00——164,000 ⚠164,000
SiliconFlow (China)
Pro/deepseek-ai/DeepSeek-V3
Cloud$0.25$1.00——164,000 ⚠164,000
Deep Infra
deepseek-ai/DeepSeek-V3
Cloud$0.32$0.89——163,840 ⚠8,192
Eden AI
deepinfra/deepseek-ai/DeepSeek-V3
Gateway$0.32$0.89——163,840 ⚠8,192
Merge GatewayGateway$0.27$1.10——163,840 ⚠81,920
Pioneer
deepseek-ai/DeepSeek-V3
Gateway$0.27$1.12$0.14$0.27163,840 ⚠8,192
D.Run (China)Gateway$0.28$1.10——131,0728,192
Alibaba (China)First-party$0.29$1.15——65,536 ⚠8,192

Sorted by blended price (input×0.75 + output×0.25) asc. The official channel always shows regardless of rank. Whether a gateway's low price is actually usable can't be verified.

Your usage cost

1Pioneer$93.80
2SiliconFlow$100.00
3SiliconFlow (China)$100.00
4SiliconFlow (China)$100.00
5Deep Infra$108.50
6Eden AI$108.50

1 more channels offer $0 (iFlow); free tiers usually have rate limits and no SLA, excluded from ranking.

The cheapest paid channel is the only channel.

Price historyone sample accumulated per data sync

Input listOutput listMin blended
$1.25$02026-08-052026-09-032026-08-05 · Input list $1.252026-08-06 · Input list $1.252026-08-07 · Input list $1.252026-08-08 · Input list $1.252026-08-09 · Input list $1.252026-08-10 · Input list $1.252026-08-11 · Input list $1.252026-08-12 · Input list $1.252026-08-13 · Input list $1.252026-09-03 · Input list $1.252026-08-05 · Output list $1.252026-08-06 · Output list $1.252026-08-07 · Output list $1.252026-08-08 · Output list $1.252026-08-09 · Output list $1.252026-08-10 · Output list $1.252026-08-11 · Output list $1.252026-08-12 · Output list $1.252026-08-13 · Output list $1.252026-09-03 · Output list $1.252026-08-05 · Min blended $0.492026-08-06 · Min blended $0.492026-08-07 · Min blended $0.492026-08-08 · Min blended $0.492026-08-09 · Min blended $0.462026-08-10 · Min blended $0.462026-08-11 · Min blended $0.462026-08-12 · Min blended $0.462026-08-13 · Min blended $0.462026-09-03 · Min blended $0.44

Artificial Analysis evaluations14 items

GPQA Diamond55.7%
Humanity's Last Exam2.9%
MMLU-Pro75.2%
SciCode35.8%
LiveCodeBench35.9%
Terminal-Bench Hard6.8%
Terminal-Bench 2.116.9%
AIME 202526.0%
AIME25.3%
MATH-50088.7%
τ²-Bench Telecom22.8%
τ³-Bench Banking4.7%
AA-LCR29.3%
IFBench34.8%

Individual evaluations run by Artificial Analysis, on the same reasoning tier as the intelligence score above. Each benchmark has its own task set and harness, so rows are not comparable with one another. The Intelligence Index above draws on a different, newer set of evaluations.

Related models

DeepSeek: DeepSeek Pro Latestsame series$0.19 / $4.20DeepSeek V4 Flash Latestsame series$0.015 / $1.28DeepSeek Latestsame series$1.10 / $2.50Qwen3 Coder Flashcheaper alternative$0.30 / $1.50Command Rcheaper alternative$0.15 / $0.60Gemma 3 12B ITcheaper alternative$0.27 / $0.27Qwen3 30B A3B Instruct 2507cheaper alternative$0.30 / $0.50
Data partly from models.dev (MIT) · Together AI official docs ↗