DeepSeek V4 Flash 0731 (batch)
by DeepSeek · deepseek/deepseek-v4-flash-0731:batch · #33 cheapest of 457 paid models
Prices updated Nov 18, 2026, 4:47 PM UTC · refreshed nightly
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....
Compare vs:Claude Sonnet 5Claude Opus 4.6Claude Haiku 4.5GPT-5.6 Sol
Pricing
Input
$0.048
per 1M tokens
Output
$0.144
per 1M tokens
Blended (3:1)
$0.072
3 input : 1 output
Cache read
$0.012
per 1M cached tokens
Cache write
—
per 1M tokens
Cache & batch economics
Effective prices for DeepSeek V4 Flash 0731 (batch) with prompt caching.
| Cache hit rate | Effective blended $/1M | Example request* | vs no cache |
|---|---|---|---|
| 0% | $0.072 | $0.00504 | — |
| 50% | $0.058 | $0.00324 | −36% |
| 90% | $0.048 | $0.0018 | −64% |
*Example: 100K sticky context + 2K new input + 1K output. Batch discount is an approximate 50% off list rates (available for OpenAI / Anthropic / Google — not flagged for this creator). Confirm on the vendor’s pricing page.
What a request costs
| Workload | Input tokens | Output tokens | Cost / request | Cost / 1K requests |
|---|---|---|---|---|
| Short chat message | 500 | 300 | $6.72e-5 | $0.0672 |
| Document summary | 8,000 | 1,000 | $0.000528 | $0.528 |
| Codebase question (RAG) | 30,000 | 2,000 | $0.001728 | $1.73 |
| Long-context analysis | 150,000 | 5,000 | $0.00792 | $7.92 |
Performance
Intelligence Index
—
AA composite quality
Coding Index
69.1
Math Index
—
Agentic Index
41
Output speed
—
median
Time to first token
—
median
Time to first answer
—
after reasoning tokens
Design Arena Elo
| Category | Elo |
|---|---|
| Website | 1,229 |
| UI components | 1,225 |
| Data viz | 1,169 |
| SVG | 1,169 |
| 3D | 1,188 |
| Game dev | 1,197 |
| Code categories | 1,217 |
| ASCII art | 1,082 |
Specs
Context window
1.05M
tokens
Max output
384K
tokens
Input modalities
text
Tokenizer
DeepSeek