DeepSeek pricing: models and monthly cost
Compare both DeepSeek V4 models by input, cache-hit and output rates. DeepSeek publishes a peak rate and an off-peak (valley) rate; Code By AI prices the valley rate all day.
Every number on this page uses DeepSeek's off-peak (valley) list rate around the clock, with no peak surcharge. The list-price column, savings percentage and monthly example are calculated from those valley rates. Neither model is a channel default: send deepseek-v4.1-flash or deepseek-v4-pro explicitly.
What a working month costs
Computed on deepseek-v4.1-flash. The example workload is 20M input and 2M output tokens; official and Code By AI prices use the same tokens.
| Usage tier | Input / output tokens | List price | Code By AI | Saving |
|---|---|---|---|---|
| Light | 5M / 500K | $1.05 | $0.32 | 70% |
| Regular | 20M / 2M | $4.20 | $1.26 | 70% |
| Heavy | 80M / 8M | $16.80 | $5.04 | 70% |
Set up DeepSeek
Export OPENAI_BASE_URL=https://www.codebyai.net/api/agent and OPENAI_API_KEY, then send OpenAI-compatible Chat Completions requests with an explicit model ID.
export OPENAI_BASE_URL="https://www.codebyai.net/api/agent"
export OPENAI_API_KEY="sk-proj-xxxx"
# OpenAI-compatible Chat Completions; send one explicit model ID:
# deepseek-v4.1-flash or deepseek-v4-pro
curl "$OPENAI_BASE_URL/v1/chat/completions" \
-H "Authorization: Bearer $OPENAI_API_KEY" \
-H "content-type: application/json" \
-d '{"model":"deepseek-v4.1-flash","messages":[{"role":"user","content":"Hello"}]}'DeepSeek models and token rates
| Model | List price | You pay |
|---|---|---|
| deepseek-v4.1-flash | $0.15 / $0.003 / $0.60 | $0.045 / $0.0009 / $0.18 |
| deepseek-v4-pro | $0.66 / $0.022 / $1.98 | $0.198 / $0.0066 / $0.594 |
per 1M tokens (input / cache hit / output)
FAQ
- How much does deepseek-v4.1-flash cost per month?
- For the example month, deepseek-v4.1-flash costs $4.20 at list price and $1.26 through Code By AI.
- Can I use DeepSeek with the same key?
- Yes. One Code By AI key works for DeepSeek and the other channels. Use the protocol and model ID shown in the guide.
- Where do I set up DeepSeek?
- Use the DeepSeek setup guide linked on this page. It covers the exact config file and model IDs for DeepSeek.