DeepSeek V3 (Dec '24) API Pricing

DeepSeek · released 2024-12-26

Pricing

$0.32
input / 1M tokens
0.032¢ / 1K
$0.89
output / 1M tokens
0.089¢ / 1K
$0.463
blended / 1M tokens
0.046¢ / 1K

What it actually costs

WorkloadTokensCost
Short chat reply500 in / 300 out0.0427¢
RAG query4,000 in / 500 out0.173¢
Long document summary50,000 in / 1,000 out$0.0169
Code review of a 2k-line file30,000 in / 2,000 out$0.0114
1,000 chat replies/day for 30 days30,000 × (500 in / 300 out)$12.81

Cost = (input tokens ÷ 1,000,000 × $0.32) + (output tokens ÷ 1,000,000 × $0.89).

Benchmarks

8.5
Intelligence
#305 of 533
23.0
Coding
#135 of 197
26.0
Math
#175 of 249
Tokens/sec
Time to first token

Similar intelligence, lower price

ModelIntelligenceBlended $/1M
Llama 3.1 Instruct 8B6.9$0.028
Gemma 4 E4B (Reasoning)8.9$0.04
Sarvam 30B (high)6.6$0.047
Granite 4.2 3B9.1$0.052
Nova Micro5.9$0.061

Leaderboard neighbours

ModelIntelligenceBlended $/1M
Devstral 28.6
Magistral Small 1.28.6$0.75
Qwen3 32B (Reasoning)8.6$0.28
DeepSeek V3 (Dec '24)8.5$0.463
Gemini 2.5 Flash-Lite (Reasoning)8.5$0.175
DeepSeek R1 Distill Qwen 32B8.4
GLM-4.6V (Non-reasoning)8.4$0.45

More from DeepSeek

ModelIntelligenceBlended $/1M
DeepSeek V4.1 Flash (Reasoning, Max Effort)39.5$0.525
DeepSeek V4 Pro 0813 (Reasoning, Max Effort)36.0$1.98
DeepSeek V4 Flash Vision (Reasoning, Max Effort)34.8$0.66
DeepSeek V4 Flash 0731 (Reasoning, Max Effort)34.3$0.66
DeepSeek V4 Pro 0424 (Reasoning, Max Effort)30.4$0.544
DeepSeek V4 Flash 0420 (Reasoning, High Effort)26.0$0.168
DeepSeek V3.2 (Reasoning)21.5$0.315
DeepSeek V3.2 Exp (Reasoning)16.6$0.315

Count tokens & estimate cost for DeepSeek V3 (Dec '24) See DeepSeek V3 (Dec '24) on the leaderboard

Source: Artificial Analysis · last synced 2026-09-22

FAQ

Who develops DeepSeek V3 (Dec '24)?

DeepSeek V3 (Dec '24) is developed by DeepSeek, released 2024-12-26.

How much does DeepSeek V3 (Dec '24) cost per million tokens?

DeepSeek V3 (Dec '24) costs $0.32 per 1M input tokens and $0.89 per 1M output tokens ($0.463 blended), per Artificial Analysis's official pricing data.

How does DeepSeek V3 (Dec '24) rank on benchmarks?

DeepSeek V3 (Dec '24) scores 8.5 on the Artificial Analysis Intelligence Index, ranking #305 of 533 models we track.

Is there a cheaper model with similar intelligence to DeepSeek V3 (Dec '24)?

Yes — Llama 3.1 Instruct 8B scores similar intelligence (6.9) at $0.028/1M blended, versus $0.463/1M for DeepSeek V3 (Dec '24).

Where does this data come from?

Pricing and benchmark data for DeepSeek V3 (Dec '24) comes from Artificial Analysis (artificialanalysis.ai), synced daily. Last synced 2026-09-22.