specification: API Commons Plans
specificationVersion: '0.1'
schema: https://raw.githubusercontent.com/api-evangelist/interface-research/main/schema/api-commons.yml#/$defs/Plans
provider: DeepSeek
providerId: deepseek
created: '2026-05-04'
# Provenance stamped 2026-08-11: this artifact was written by the API Evangelist
# bulk sweep dated 2026-05-04, not harvested from the provider. See roadmap#35.
method: generated
modified: '2026-05-05'
reconciled: true
tags:
  - AI
  - Artificial Intelligence
  - Chat
  - LLM
  - Large Language Models
  - Token-Based Pricing
description: DeepSeek prices its API on a per-token, pay-as-you-go basis with no monthly subscription
  tier. Two production models are available — DeepSeek-V4-Flash (non-thinking) and DeepSeek-V4-Pro
  (thinking/reasoning). Both expose distinct prices for cache-hit and cache-miss input tokens plus output
  tokens. The legacy `deepseek-chat` and `deepseek-reasoner` model identifiers are deprecated aliases.
  V4-Pro carries a 75 percent promotional discount valid through 2026-05-31. Account top-ups are prepaid;
  there is no separate enterprise SKU on the public site.
sources:
  - https://api-docs.deepseek.com/quick_start/pricing
  - https://api-docs.deepseek.com/quick_start/rate_limit
plans:
  - id: deepseek-v4-flash
    name: DeepSeek-V4-Flash
    type: usage-based
    description: Pay-as-you-go pricing for DeepSeek-V4-Flash (non-thinking mode), the default chat
      and FIM completion model. Replaces the deprecated `deepseek-chat` alias.
    entries:
      - label: Input tokens (cache hit)
        name: input_tokens_cache_hit
        type: metered
        metric: tokens
        limit: -1
        timeFrame: usage
        geo: global
        unit: 1000000
        price: '$0.0028'
        userMultiplied: false
      - label: Input tokens (cache miss)
        name: input_tokens_cache_miss
        type: metered
        metric: tokens
        limit: -1
        timeFrame: usage
        geo: global
        unit: 1000000
        price: '$0.14'
        userMultiplied: false
      - label: Output tokens
        name: output_tokens
        type: metered
        metric: tokens
        limit: -1
        timeFrame: usage
        geo: global
        unit: 1000000
        price: '$0.28'
        userMultiplied: false
    elements:
      - name: 1M token context window
      - name: 384K max output tokens
      - name: Chat Completions API
      - name: Fill-In-The-Middle (FIM) Completions API
      - name: JSON output
      - name: Tool calls
      - name: Chat prefix completion
      - name: Context caching
  - id: deepseek-v4-pro
    name: DeepSeek-V4-Pro
    type: usage-based
    description: Pay-as-you-go pricing for DeepSeek-V4-Pro (thinking / reasoning mode). Replaces
      the deprecated `deepseek-reasoner` alias. Listed prices reflect the 75 percent promotional
      discount in effect through 2026-05-31.
    entries:
      - label: Input tokens (cache hit)
        name: input_tokens_cache_hit
        type: metered
        metric: tokens
        limit: -1
        timeFrame: usage
        geo: global
        unit: 1000000
        price: '$0.003625'
        userMultiplied: false
      - label: Input tokens (cache miss)
        name: input_tokens_cache_miss
        type: metered
        metric: tokens
        limit: -1
        timeFrame: usage
        geo: global
        unit: 1000000
        price: '$0.435'
        userMultiplied: false
      - label: Output tokens
        name: output_tokens
        type: metered
        metric: tokens
        limit: -1
        timeFrame: usage
        geo: global
        unit: 1000000
        price: '$0.87'
        userMultiplied: false
    elements:
      - name: 1M token context window
      - name: 384K max output tokens
      - name: Reasoning / thinking mode
      - name: Chat Completions API
      - name: JSON output
      - name: Tool calls
      - name: Chat prefix completion
      - name: Context caching
      - name: 75% promotional discount through 2026-05-31
maintainers:
  - FN: Kin Lane
    email: kin@apievangelist.com
x-canonical:
  generated: '2026-10-08'
  vocabulary: all/0-working/economics/plans-canonical.yml
  models:
  - metered
  units:
  - token
  plans:
  - name: DeepSeek-V4-Flash
    access: other
    model: metered
    model_how: stated
    units:
    - token
    periods:
    - per-use
  - name: DeepSeek-V4-Pro
    access: other
    model: metered
    model_how: stated
    units:
    - token
    periods:
    - per-use