Price checkMonthly
Granted Balance (Trial Credits)CustomDeepSeek V4-Flash (Pay-as-you-go)CustomDeepSeek V4-Pro (Pay-as-you-go)Custom
See all 3 plans

All DeepSeek API Plans & Pricing

Prices unverified — last checked Oct 20, 2026

Plan Monthly Annual Best For
View all features by plan (compare side-by-side)

Granted Balance (Trial Credits)

  • Granted balance on signup (check platform for current amount)
  • No credit card required
  • Access to all models
  • Credits apply automatically to usage

DeepSeek V4-Flash (Pay-as-you-go)

  • $0.15/M input tokens (cache miss, off-peak) / $0.30 peak
  • $0.003/M input tokens (cache hit, off-peak) / $0.006 peak
  • $0.60/M output tokens (off-peak) / $1.20 peak
  • Off-peak = all hours except Mon–Fri 01:00–04:00 and 06:00–10:00 UTC
  • Model: deepseek-flash (DeepSeek-V4.1-Flash-1017); non-thinking and thinking modes
  • 1M context length, 384K max output
  • Vision input supported
  • 2500 concurrent request limit

DeepSeek V4-Pro (Pay-as-you-go)

  • $0.66/M input tokens (cache miss, off-peak) / $1.32 peak
  • $0.022/M input tokens (cache hit, off-peak) / $0.044 peak
  • $1.98/M output tokens (off-peak) / $3.96 peak
  • Off-peak = all hours except Mon–Fri 01:00–04:00 and 06:00–10:00 UTC
  • Higher capability flagship model
  • 500 concurrent request limit
Pricing Alerts

Track DeepSeek API pricing

Get an email when DeepSeek API's pricing changes — plus the weekly SaaS Price Watch: verified price changes and deals across 3,000+ products. One-click unsubscribe.

Request DeepSeek API Pricing
Compare DeepSeek API with alternativesAdjust seats, lock a tier, add up to 2 more products side-by-side. Shareable URL.
Quick Answer
Last verified:
Medium confidence

DeepSeek API uses custom pricing as of October 2026 with 3 plans available. Contact DeepSeek API directly for a personalized quote. Pricing depends on your chosen tier, contract length, and negotiated discounts.

Use the interactive pricing calculator to estimate your exact cost based on team size and requirements.

  • Free tier: No free tier available

DeepSeek API offers 3 pricing tiers: Granted Balance (Trial Credits), DeepSeek V4-Flash (Pay-as-you-go), DeepSeek V4-Pro (Pay-as-you-go). The DeepSeek V4-Flash (Pay-as-you-go) plan is most code generation tasks at low per-token cost.

Compared to other code generation apis software, DeepSeek API is positioned at the budget-friendly price point.

  • 3 documented hidden costs beyond list price
Last verified: | Sources: 2 pricing pages and reviews

How much does DeepSeek API cost?

DeepSeek API uses custom pricing across 3 plans. Contact DeepSeek API directly for a personalized quote. Plans include Granted Balance (Trial Credits) (custom pricing), DeepSeek V4-Flash (Pay-as-you-go) (custom pricing), DeepSeek V4-Pro (Pay-as-you-go) (custom pricing).

DeepSeek API Pricing Overview

DeepSeek API uses custom pricing — contact their sales team for a quote. The Granted Balance (Trial Credits) plan requires contacting sales for a custom quote and is designed for new users evaluating the api before topping up a balance. The DeepSeek V4-Flash (Pay-as-you-go) plan requires contacting sales for a custom quote and is designed for most code generation tasks at low per-token cost. The DeepSeek V4-Pro (Pay-as-you-go) plan requires contacting sales for a custom quote and is designed for complex reasoning and flagship-level coding performance.

DeepSeek API with a No minimum spend for individual developers. minimum commitment.

There are at least 3 documented hidden costs beyond DeepSeek API's list price, including implementation, training, and add-on fees.

This pricing was last verified in October 20, 2026 from 2 independent sources.

Request DeepSeek API Pricing

DeepSeek API provides access to DeepSeek's code-optimized and reasoning models via a pay-per-token API compatible with the OpenAI and Anthropic SDK formats. The current models are deepseek-flash (DeepSeek V4.1 Flash, fast and low-cost) and DeepSeek V4-Pro (higher capability), replacing the legacy deepseek-chat and deepseek-coder endpoints. Pricing is among the most competitive in the code generation API space: deepseek-flash input tokens cost $0.30 per million at peak and $0.15 off-peak. Context caching cuts cached input to about 2–3% of the cache-miss price. Usage is deducted from a prepaid balance, with any granted balance used first.

How DeepSeek API Pricing Compares

Compare DeepSeek API pricing against top alternatives in Code Generation APIs.

Codestral API $0.2-$0.9 per million tokens Compare → Refact.ai $0-$10 per user per month Compare → Tabby $0-$19 per user per month Compare →

Compare DeepSeek API vs Alternatives

Before committing to DeepSeek API, compare pricing with these 3 alternatives in the same category.

VSCodestral API
From $0.2 per million tokens

Teams embedding code-gen into editors, CI tools, or internal platforms

Full comparison
VSRefact.ai
Free

Individual devs trying the cloud product or self-hosting

Full comparison
VSTabby
Free

Hobbyists and small teams self-hosting

Full comparison

All DeepSeek API alternatives & migration guides

What Companies Actually Pay for DeepSeek API

Review scores
Third-party review aggregates, as of Oct 2026

How DeepSeek API Pricing Compares

Software Starting Price Top Price
DeepSeek API Custom Custom
Codestral API $0.3 per million tokens $0.9 per million tokens
Refact.ai Free $10 per user per month
Tabby Free $19 per user per month
Augment Code API $20/month $100/month
Codeium API Free $200/month

Detailed pricing comparisons:

Browse all Code Generation APIs pricing →

3 DeepSeek API Hidden Costs Beyond the List Price

Beyond the listed price, DeepSeek API has at least 3 documented hidden costs that can significantly increase total cost of ownership.

Watch for 3 hidden costs
  • Thinking Mode Billing
    medium 1 source
    Grounded summary These reasoning tokens are billed at the output rate, which can increase the total cost even if the visible output is short
  • Data Exposure to Chinese Regimes
    critical 1 source
    Grounded summary Data Residency and Security: For enterprises, using the hosted DeepSeek V4 API means accepting data exposure to Chinese legal and regulatory regimes, which can be a "hard blocker" for workloads involving personally identifiable information (PII), ...
  • Content Restrictions
    high 1 source
    Grounded summary Content Restrictions: DeepSeek V4 has documented limitations on topics related to Chinese politics and government, which could lead to incomplete or evasive outputs for applications involving global news analysis or geopolitical research
Tip

Ask your DeepSeek API sales rep about these costs upfront. Getting them in writing before signing can save you from surprise charges later.

Full hidden costs breakdown →

Intelligence sourced from 1 independent sources
industry
Key claims include inline source attribution. Data verified against multiple independent sources. 4 source citations total.

DeepSeek API Contract Terms

DeepSeek API contracts do not auto-renew. Changes require advance notice. These terms are sourced from verified buyer experiences.

Contract Terms
Minimum Commitment No minimum spend for individual developers.

How to Negotiate DeepSeek API Pricing

DeepSeek API contracts are negotiable. These 1 tactics are sourced from real buyer experiences and procurement specialists.

Negotiation Playbook 1 tactics
Negotiate Enterprise Pricing high success

Pricing for enterprise needs is 'negotiable and tailored to business requirements'.

https://chat-deep.ai
High success Medium Low / difficult

Full negotiation guide →

DeepSeek API Price History

Pricing changes CostBench has tracked for DeepSeek API, verified across 2 snapshots going back to 2026Q2.

  1. DeepSeek API tier restructure

    Changed from 3 to 4 pricing tiers

View DeepSeek API's full price history →

DeepSeek API Pricing FAQ

01 How much does DeepSeek API cost?

DeepSeek's API uses pay-per-token pricing with peak and off-peak rates (off-peak is half). deepseek-flash (DeepSeek V4.1 Flash) costs $0.30 per million input tokens and $1.20 per million output tokens at peak (cache-miss rates), or $0.15/$0.60 off-peak. DeepSeek V4-Pro costs $1.32/M input and $3.96/M output at peak, or $0.66/$1.98 off-peak.

02 Does DeepSeek offer a free tier for the API?

There is no ongoing free tier. DeepSeek's pricing page says usage is deducted from your topped-up balance or a granted balance (the granted balance is used first when both exist), but it does not publish a standard free-credit amount for new accounts. After any granted balance is used, usage is billed at the pay-per-token rates.

03 What happened to the deepseek-coder model endpoint?

DeepSeek's current API model names are deepseek-flash and deepseek-v4-pro. Its pricing page says the legacy names deepseek-v4-flash and deepseek-v4-flash-vision-exp are still accepted but their models have been retired: those requests are served by DeepSeek-V4.1-Flash-1017 and billed at the Flash price. The older deepseek-coder name is not listed, so developers should call deepseek-flash or deepseek-v4-pro directly.

04 How is DeepSeek API billed?

DeepSeek API is billed on a pay-as-you-go basis. The available paid tiers are DeepSeek V4-Flash (Pay-as-you-go) and DeepSeek V4-Pro (Pay-as-you-go); vision input is now handled by the Flash model at the Flash price. There are no fixed monthly subscription fees — you pay only for what you use.

05 What happens when my granted balance runs out?

Once any granted balance is used up or expires, you will need to top up your balance to keep using the DeepSeek API. There is no ongoing free tier.

Is this pricing incorrect? — we'll verify and update it.