DeepSeek-V4 hybrid model with both non-thinking and thinking (default) modes.
| Date | Input | Output | Cache read | Change |
|---|---|---|---|---|
| Sep 2, 2026 | $0.22 | $0.66 | $0.007 | — |
| Sep 26, 20269cb03f8 | $1 | $1 | $1 | ↑355% |
Input is ↑355% since Sep 2, 2026, from $0.22 to $1 per 1M tokens.
| Provider | Input | Output | Since launch |
|---|---|---|---|
| DeepInfracheapest | $0.44 | $1.32 | — |
| Fireworks AI | $1 | $1 | ↑355% |
| Model | Context | Max out | Pricing |
|---|---|---|---|
| deepseek-v4-flash-0731 | 1.0M | — | $0.14$0.28 |
| deepseek-v4-flash | 1.0M | 384K | $0.14$0.28 |
| deepseek-v4p1-flash | 1.0M | — | $0.3$1.2 |
| deepseek-v3p1-terminus | 163K | — | $1$1 |
| deepseek-v4-flash-vision-exp | 1.0M | 384K | $1$1 |
| deepseek-coder-1b-base | 16K | — | $1$1 |
| deepseek-coder-33b-instruct | 16K | — | $1$1 |
| deepseek-coder-7b-base-v1p5 | 4K | — | $1$1 |
| deepseek-coder-7b-base | 4K | — | $1$1 |
| deepseek-coder-7b-instruct-v1p5 | 4K | — | $1$1 |
| deepseek-coder-v2-instruct | 33K | — | $1$1 |
| deepseek-coder-v2-lite-base | 163K | — | $1$1 |
| deepseek-coder-v2-lite-instruct | 163K | — | $1$1 |
| deepseek-prover-v2 | 163K | — | $1$1 |
| deepseek-v2-lite-chat | 163K | — | $1$1 |
| deepseek-v2p5 | 33K | — | $1$1 |
| deepseek-v3-0324 | 163K | — | $1$1 |
| deepseek-v3 | 131K | — | $1$1 |
| deepseek-v3p1 | 163K | — | $1$1 |
| deepseek-v3p2 | 163K | — | $1$1 |
| deepseek-v4-flash | 1.0M | 384K | $1$1 |
| deepseek-v4-pro | 1.0M | 384K | $1$1 |
| deepseek-v4-pro-0813 | 1.0M | — | $1.32$3.96 |
| deepseek-v4-pro | 1.0M | 384K | $1.74$3.48 |
/v1/models/fireworks/deepseek-ai/deepseek-v4-flash-vision-exp