DeepSeek V4 Flash 0731
DeepSeek V4 Flash 0731 is a sparse mixture-of-experts model from DeepSeek, with 13B active parameters out of 284B total. This re-post-trained revision is suited for coding, reasoning, and agent workflows.
Cheapest input
$0.09
$/MTok · OpenRouter
Cheapest output
$0.18
$/MTok
Context
1.0M
tokens
Offers
3
2 sources
Price changes
Cheapest list input/output only. Not a daily candle chart: rates move rarely, so this is an event log. Arrow down means cheaper.
- ↓ dropOutput $0.28 → $0.18
- ↓ dropInput $0.14 → $0.09
Offers
Same model, different listings. First-party, router, and aggregator rows stay separate so you can see the spread instead of a single blended number.
| Source | Provider | Input | Output | Cache read | Context |
|---|---|---|---|---|---|
| OpenRoutercheapestdeepseek/deepseek-v4-flash-0731 | OpenRouter | $0.09 | $0.18 | $0.018 | 1.0M |
| models.devdeepseek/deepseek-v4-flash-0731 | Vercel AI Gateway | $0.13 | $0.26 | $0.028 | 1M |
| models.devdeepseek/deepseek-v4-flash-0731 | Kilo Gateway | $0.14 | $0.28 | $0.028 | 1.0M |
Capabilities
- Reasoning
- Tool calling
- Structured output
- Open weights
- In: text
- Out: text
Knowledge cutoff: 2025-05Released: 2026-07-31Family: deepseek-flash
Scores
Sparse coverage on purpose. Missing a score means we have not linked one yet, not that the model is unranked. Full boards on /benchmarks.
- AA intelligence49.9
- AA coding69.1
- AA agentic45.7
Sources: Arena AI, Artificial Analysis. See /benchmarks.