DeepSeek-V4-Flash (2026-07-31)
- In Input modalities: Text Out Output modalities: Text
- Reasoning· low, high, max
- Tool use
- Structured output
- Open weights
Prices
9 offerings · prices as listeddeepseek-v4-flash-0731| Conditions/ 1M tokens | Input | Output |
|---|---|---|
| Peak hours Standard | $0.424 | $1.272 |
| Off-peak hours | $0.212 | $0.636 |
Prices from Alibaba Cloud Model Studio pricing (opens in a new tab) (checked Nov 2, 2026)
deepseek-v4-flash-0731| Conditions/ 1M tokens | Input | Output |
|---|---|---|
| Standard | ¥3.00 | ¥9.00 |
| Off-peak hours | ¥1.50 | ¥4.50 |
Prices from 阿里云百炼：模型价格 (opens in a new tab) (checked Nov 2, 2026)
deepseek-v4-flash-0731| Conditions/ 1M tokens | Input | Output |
|---|---|---|
| Peak hours Standard | $0.44 | $1.32 |
| Off-peak hours | $0.22 | $0.66 |
Prices from Alibaba Cloud Model Studio pricing (opens in a new tab) (checked Nov 2, 2026)
deepseek-v4-flash-0731| Conditions/ 1M tokens | Input | Output |
|---|---|---|
| Peak hours no longer listed since Oct 23, 2026 | $0.424 | $1.272 |
| Off-peak hours no longer listed since Oct 23, 2026 | $0.212 | $0.636 |
Prices from Alibaba Cloud Model Studio pricing (opens in a new tab) (checked Nov 2, 2026)
Standard: the price a request with no conditions pays (the headline price elsewhere on the site).
Details
- Knowledge cutoff
- May 2025
- Licence
- MIT
- Family
- deepseek-flash
- Model ID
Not stated on DeepSeek’s pages: max input, reasoning can be off, max thinking budget, built-in tools, aliases.
Run locally
Weights: deepseek-ai/DeepSeek-V4-Flash-0731- Parameters 304.2B
- Active 33.6B (estimated) · 256 experts, 6 per token
- Native precision FP8
- Layers 43
- KV cache per token 86 KB
- Config positions 1024K (config.json max_position_embeddings)
- API context 977K (maker’s API)
- Needs
- ≈ 158.7 GB (weights 150.5 + KV cache 0.7 + overhead 7.5)
- Usable
- 21.6 GB (90% of its memory)
- Try
- IQ2_XXS runs (needs system memory), but below 4 bits quality drops noticeably
| Precision | Weights | Needs at 8K | On your hardware | Longest context |
|---|---|---|---|---|
| FP8 (native)maker’s file | 155.4 GB | ≈ 163.9 | ✕Doesn’t fit | — |
| Q6_K | 232.7 GB | ≈ 245.0 | ✕Doesn’t fit | — |
| Q5_K_M | 201.5 GB | ≈ 212.2 | ✕Doesn’t fit | — |
| Q5_0 | 194.8 GB | ≈ 205.2 | ✕Doesn’t fit | — |
| Q4_K_M | 172.1 GB | ≈ 181.4 | ✕Doesn’t fit | — |
| Q4_0 | 159.4 GB | ≈ 168.0 | ✕Doesn’t fit | — |
| IQ4_XS | 150.5 GB | ≈ 158.7 | ✕Doesn’t fit | — |
| Q3_K_Mquality drops | 138.5 GB | ≈ 146.1 | ✕Doesn’t fit | — |
| IQ3_XXSquality drops | 108.4 GB | ≈ 114.4 | ✕Doesn’t fit | — |
| Q2_Kquality drops | 106.2 GB | ≈ 112.2 | ✕Doesn’t fit | — |
| IQ2_XXSquality drops | 72.9 GB | ≈ 77.3 | ○Needs system memory 7–10 tok/s | — |
Click a row to choose a precision. “Maker’s file” sizes are the maker’s own files; others are parameters × bits per weight.
Compared with other hardware
Each at the highest precision it runs. Click one for details.
Smallest that runs it (Q4_K_M, 8K): no single listed GPU; unified memory Apple M5 Ultra (64-core GPU) · 256 GB.
Run commands
The maker publishes no files at this precision. Choose its native precision or one of its own builds, or quantise it yourself with llama.cpp.
Memory to fine-tune
Batch 1, 2,048-token sequences, gradient checkpointing; LoRA and QLoRA train rank-16 adapters on the attention projections.
Estimates; facts from the maker’s Hugging Face repositories, as of 2026-11-06. How it’s estimated
Cited scores
From third-party leaderboards, not measured by us- Epoch Capabilities Index 154.3
#48 of 118 listed models · CI 152.6–156.3
- Epoch AI · Chess Puzzles 32%
#47 of 85 listed models · max effort · ±4.69
- Epoch AI · FrontierMath Tier 4 v2 24.4%
#42 of 65 listed models · max effort · ±6.79
- Epoch AI · FrontierMath Tiers 1–3 57.5%
#43 of 84 listed models · max effort · ±2.93
- Epoch AI, 'Capabilities & benchmarking'. Published online at epoch.ai. Retrieved from 'https://epoch.ai/benchmarks'. Licensed CC BY 4.0. · License · as of Oct 28 – Nov 5, 2026
- EQ-Bench by Samuel J. Paech (eqbench.com; EQ-bench/EQ-bench-site on GitHub), MIT License. · License · as of Oct 28, 2026
- LiveBench (White et al.), release 2026-06-25, livebench.ai, licensed CC BY-SA 4.0. This board, adapted from it, is shared under CC BY-SA 4.0. · License · as of Jun 25, 2026
- changes we made
Changes
All changes →- New language model
- Removed from the directory
Sources
- Cloudflare Workers AI pricing (opens in a new tab) developers.cloudflare.com · checked Oct 31, 2026
- Alibaba Cloud Model Studio pricing (opens in a new tab) alibabacloud.com · checked Nov 2, 2026
- deepseek-ai/DeepSeek-V4-Flash-0731 model card (opens in a new tab) huggingface.co · checked Oct 30, 2026
- 阿里云百炼：模型价格in Chinese (opens in a new tab) help.aliyun.com · checked Nov 2, 2026
- Explore Qwen3, Wan2.6 & Latest Models on Model Studio - Alibaba Cloud Model Studio - Alibaba Cloud_Alibaba Cloud Model Studio(Model Studio) - Alibaba Cloud Model Studio - Alibaba Cloud Documentation Center (opens in a new tab) alibabacloud.com · checked Nov 2, 2026
- DeepSeek API change log (opens in a new tab) api-docs.deepseek.com · checked Nov 2, 2026
- models.dev: deepseek/deepseek-v4-flash-0731 (community data, MIT) (opens in a new tab) github.com · checked Oct 11, 2026
- 模型上下架与更新-大模型服务平台百炼-阿里云_大模型服务平台百炼(Model Studio)-大模型服务平台百炼(Model Studio)-阿里云帮助中心in Chinese (opens in a new tab) help.aliyun.com · checked Nov 4, 2026
- DeepSeek (opens in a new tab) alibabacloud.com · checked Nov 1, 2026
- deepseek-ai/DeepSeek-V4-Flash-0731 model card (opens in a new tab) huggingface.co · checked Nov 1, 2026
- deepseek-ai/DeepSeek-V4-Flash-0731 config.json (opens in a new tab) huggingface.co · checked Nov 3, 2026
- deepseek-ai/DeepSeek-V4-Flash-0731 files (opens in a new tab) huggingface.co · checked Nov 3, 2026
DeepSeek’s official policies: data use, privacy, terms →
Some values on this page are still waiting to be checked against official pages: a filled dot is verified, a hollow one is not yet (with or without a source on file). How we verify data
DeepSeek · China · deepseek.com
Related models
Snapshots, the same family and other models from this provider