DeepSeek-V3 pricing & specs
DeepSeek · released 2024-12-26 · ● verified 2026-10-11
input / 1M$0.32via DeepInfra
output / 1M$0.89via DeepInfra
context163.8K
max output8.2K
// price history ($/1M tokens)
// cheaper or similar alternatives
| model | in | out | ctx | vs DeepSeek-V3 |
|---|---|---|---|---|
| Llama 3.2 1B Instruct | $0.02 | $0.02 | 131K | −95.7% |
| Gemma 4 E2B IT | $0.04 | $0.08 | 131.1K | −89.2% |
| Qwen3.7 Flash | $0.03 | $0.13 | 1M | −88.1% |
| Gemma 3 12B IT | $0.05 | $0.10 | 131.1K | −86.5% |
| Command R7B | $0.0375 | $0.15 | 128K | −85.8% |
// price by host
| host | in $/1M | out $/1M |
|---|---|---|
| DeepInfra | $0.32 | $0.89 |
| Novita AI | $0.89 | $0.89 |
// faq
How much does DeepSeek-V3 cost?
DeepSeek-V3 costs $0.32 per million input tokens and $0.89 per million output tokens (cheapest tracked host: DeepInfra). At 50M input and 10M output tokens per month that is about $24.90.
What is the context window of DeepSeek-V3?
163.8K tokens, with up to 8.2K output tokens per request.
What are cheaper alternatives to DeepSeek-V3?
Llama 3.2 1B Instruct ($0.02 in / $0.02 out), Gemma 4 E2B IT ($0.04 in / $0.08 out), Qwen3.7 Flash ($0.03 in / $0.13 out) offer the same input types and capabilities for less.
Is DeepSeek-V3 open weights?
Yes. It is available from 2 tracked hosts; the cheapest is DeepInfra at $0.32 input and $0.89 output per million tokens.
// spec sheet
- input / 1M
- $0.32
- output / 1M
- $0.89
- blended (3:1)
- $0.4625
- context
- 163.8K
- max output
- 8.2K
- input types
- text
- reasoning
- no
- tool calling
- yes
- structured output
- yes
- open weights
- yes
- knowledge cutoff
- n/a
- price source
- DeepInfra
// cost calculator
| model | in $/1M | out $/1M | = monthly |
|---|---|---|---|
| Llama 3.2 1B Instruct | $0.02 | $0.02 | $1.20 |
| Gemma 4 E2B IT | $0.04 | $0.08 | $2.80 |
| Qwen3.7 Flash | $0.03 | $0.13 | $2.80 |
| DeepSeek-V3 | $0.32 | $0.89 | $24.90 |
Base-tier list prices. Long-context tiers, caching and batch discounts are not included.