Model comparison
Claude Haiku 4.5 vs DeepSeek-V4-Pro-0813
A spec-by-spec comparison of Claude Haiku 4.5 and DeepSeek-V4-Pro-0813: pricing, context window, capabilities, and independently sourced benchmarks.
Head-to-head
- DeepSeek-V4-Pro-0813 costs 1.5x less per input token than Claude Haiku 4.5 ($0.66/MTok vs $1.00/MTok).
- DeepSeek-V4-Pro-0813 has a 5.0x larger context window than Claude Haiku 4.5 (1,000,000 tokens vs 200,000 tokens).
- Only Claude Haiku 4.5 supports vision.
| Spec | Claude Haiku 4.5 | DeepSeek-V4-Pro-0813 |
|---|---|---|
| Provider | anthropic | deepseek |
| Family | Claude Haiku | DeepSeek V4 |
| Status | current | frontier |
| Release date | 2025-10-15 | — |
| Knowledge cutoff | 2025-02 | — |
| Context window | 200,000 tokens | 1,000,000 tokens |
| Max output tokens | 64,000 tokens | 384,000 tokens |
| Input price | $1.00/MTok | $0.66/MTok |
| Output price | $5.00/MTok | $1.98/MTok |
| Cached input price | $0.10/MTok | $0.02/MTok |
| Vision | Yes | No |
| Tool use | Yes | Yes |
| Extended thinking | Yes | Yes |
| Streaming | — | — |
Claude Haiku 4.5
Strengths- The only current Claude model that supports classic extended thinking.
- Uses the older Claude tokenizer, so token counts are ~30% lower than Sonnet 5 for the same text.
- Reliable knowledge cutoff of Feb 2025 is the oldest in the current Claude lineup.
- 200k context window against 1M on every other current Claude model.
- Does not support adaptive thinking.
DeepSeek-V4-Pro-0813
Strengths- Cache-hit input at $0.022/MTok (off-peak) is roughly 30x cheaper than this model’s own off-peak standard input rate.
- 384K max output is triple the 128K ceiling of the OpenAI and Anthropic flagships.
- Thinking and non-thinking modes are billed at one price.
- Exposes both OpenAI-format and Anthropic-format endpoints.
- Peak-hour pricing (01:00-04:00 and 06:00-10:00 UTC, Monday-Friday, excluding Chinese public holidays) doubles every rate: $1.32/MTok input, $3.96/MTok output, $0.044/MTok cached.
- Vendor label is now DeepSeek-V4-Pro-0813, a dated snapshot; the vendor page previously carried an undated label.
- Does not support image input.
- Concurrency is capped at 500 requests.
- DeepSeek states it reserves the right to adjust prices and recommends checking the page regularly.
Not sure which one to use?
Flaibo is a daily AI orientation hub: a 12-week path and a prompt vault, with Free and Professional plans. Build the habit, not just the comparison.
Start your AI path