DeepSeek-V4-Flash-0731
Released
DeepSeek-V4-Flash-0731 is an AI model released by DeepSeek on Friday, Jul 31 2026, 98 days after DeepSeek-V4-Flash. Benchmark results (shown below) cover BullshitBench v2, SWE-Bench Verified, Terminal-Bench 2.1, Toolathlon-Verified, BrowseComp, CyberGym, and 2 more.
API pricing
- InputWhat you pay for everything you send the model — your question, plus any documents or earlier conversation you include with it.
- $0.22
- Cached inputA reduced rate for text you send over and over. If every request starts with the same instructions or the same document, the provider keeps a copy ready and charges less to read it again.
- $0.007
- OutputWhat you pay for the text the model writes back. It is normally the dearer half: producing an answer costs more than reading one.
- $0.66
- InputWhat you pay for everything you send the model — your question, plus any documents or earlier conversation you include with it.
- $0.44
- Cached inputA reduced rate for text you send over and over. If every request starts with the same instructions or the same document, the provider keeps a copy ready and charges less to read it again.
- $0.014
- OutputWhat you pay for the text the model writes back. It is normally the dearer half: producing an answer costs more than reading one.
- $1.32
- Thinking and non-thinking modes use the same published price.
Available from
| OpenInference | — | $0.04 | $0.10 | 1M | fp8 |
|---|---|---|---|---|---|
| Baidu | — | $0.0352 | $0.1056 | 1M | fp8 |
| Relace | — | $0.06 | $0.12 | 1M | fp4 |
| Inceptron | — | $0.0619 | $0.1521 | 1M | fp4 |
| StreamLake | — | $0.0572 | $0.1716 | 1M | fp8 |
| DeepInfra | — | $0.06 | $0.18 | 1M | fp8 |
| Makora | — | $0.09 | $0.195 | 1M | — |
| Wafer | fast | $0.10 | $0.25 | 1M | — |
| DigitalOcean | — | $0.08 | $0.252 | 1M | — |
| BaseTen | — | $0.13 | $0.26 | 1M | fp8 |
| CoreWeave | — | $0.13 | $0.28 | 256K | fp8 |
| Parasail | — | $0.14 | $0.28 | 1M | fp8 |
| Together | — | $0.14 | $0.28 | 1M | — |
| Morph | — | $0.1234 | $0.3475 | 1M | bf16 |
| Venice | — | $0.175 | $0.35 | 1M | — |
| Sail Research | — | $0.078 | $0.36 | 1M | fp4 |
| Mancer 2 | — | $0.20 | $0.60 | 1M | fp8 |
| Reka | — | $0.11 | $0.66 | 256K | fp4 |
| Fireworks | — | $0.22 | $0.66 | 1M | — |
| GMICloud | — | $0.286 | $0.858 | 1M | fp8 |
| Alibaba | — | $0.352 | $1.056 | 1M | — |
| NextBit | — | $0.352 | $1.056 | 1M | fp8 |
| Novita | — | $0.4092 | $1.2276 | 1M | fp8 |
| AtlasCloud | — | $0.44 | $1.32 | 1M | fp4 |
| Cloudflare | — | $0.44 | $1.32 | 1.3M | — |
| Phala | — | $0.44 | $1.32 | 1M | — |
Benchmarks
Coding
Terminal & CLI
Agentic & tool use
Reasoning & science
Robustness
Source: BenchLM, retrieved 31 August 2026. Every other score here is the figure the lab published at launch.
About
DeepSeek-V4-Flash-0731, announced July 31, 2026 as the "official" V4-Flash API in public beta, was a post-training refresh of the April V4-Flash build — same architecture, retrained for agentic work. The jump was unusually large for a checkpoint update: on DeepSeek's own evaluation harness it scored 82.7 on Terminal-Bench 2.1 against 61.8 for the April build, and 76.7 on CyberGym, surpassing the larger V4-Pro-Preview on several coding benchmarks despite being the cheaper tier. DeepSeek's harness numbers for competitor models differed from those labs' own published figures, so its launch chart is best read as within-family comparison.
The release doubled as a renaming: DeepSeek retroactively designated the April 24 builds V4-Flash-Preview and V4-Pro-Preview, and served the new checkpoint under the unchanged deepseek-v4-flash API name — existing callers were switched to it without a code change. It added native support for the OpenAI Responses API format and Codex compatibility. Unlike every prior DeepSeek release, it shipped API-only at launch: no 0731 weights were published, and Hugging Face still carried the April build at release. V4-Pro, the app, and the web product stayed on their April checkpoints until the 0813 Pro refresh two weeks later. The Flash tier moved off the V4 architecture entirely with DeepSeek-V4.1-Flash in September 2026.
Get DeepSeek releases and AI news.
One email, every other Monday.
Compare DeepSeek-V4-Flash-0731 with
Suggested comparisons
Frequently asked questions
DeepSeek-V4-Flash-0731 was released by DeepSeek on Friday, Jul 31 2026.