DeepSeek AI
DeepSeek AI
DeepSeek V4 Flash and V4 Pro now define DeepSeek's hosted API lineup. Both support thinking and non-thinking modes, 1M context, tool calls, JSON output, and very low cache-hit input pricing.
chat.deepseek.comLast updated: May 5, 2026
$0.28/M
Input Price
128K
Context Window
8K
Max Output
60 t/s
Speed (3x V2)
Available Models
| Model | Input $/1M | Output $/1M | Context | Best For |
|---|---|---|---|---|
| deepseek-v4-flash (flagship) | $0.14 cache miss / $0.0028 cache hit | $0.28 | 1M | Cost-efficient reasoning, coding, and tool use |
| deepseek-v4-pro | $0.435 cache miss / $0.003625 cache hit | $0.87 | 1M | Higher-capability V4 workloads during current discount |
| deepseek-chat | Compatibility alias for V4 Flash non-thinking mode | Existing chat integrations | ||
| deepseek-reasoner | Compatibility alias for V4 Flash thinking mode | Existing reasoning integrations | ||
Note: DeepSeek offers free web and app access for consumer use. API is pay-as-you-go with no subscription required.
Strengths & Weaknesses
Strengths
- Extremely low hosted API pricing, especially on cache-hit input tokens
- Strong reasoning and coding performance at its price point
- 1M context length with tool calls and JSON output support
- OpenAI-compatible API for easy migration
- Free consumer access via web and mobile apps
Weaknesses
- Separate API vs web/app model behavior
- Data residency and governance require extra review for sensitive workloads
- V4 Pro discount is time-limited and pricing can change
- Less mature ecosystem compared to US providers
Best For
- High-volume reasoning agents at minimal cost
- Coding assistants and code review pipelines
- Cost-sensitive production workloads
- Document processing and data extraction
- Startups and developers on tight budgets
Latest Release NEW
DeepSeek V4 Flash and V4 Pro — Current API lineup
- Supports both non-thinking and thinking modes
- 1M context window with native tool call support
- Cache-hit input pricing is deeply discounted for repeated context
- V4 Flash compatibility aliases cover older deepseek-chat and deepseek-reasoner integrations
- V4 Pro has a time-limited discount in the current pricing table
Previous Releases
DeepSeek-R1 — January 2025
- Dedicated reasoning model rivaling OpenAI o1 at a fraction of the cost
- Open-source release that demonstrated competitive reasoning capabilities
- Sparked global attention for Chinese AI competitiveness
DeepSeek-V3 — December 2024
- MoE architecture trained for ~$5.5M — fraction of competitor costs
- Matched or exceeded GPT-4 level performance on key benchmarks
- Established DeepSeek as a top-tier model provider