DeepSeek V4 Flash deepseek-v4-flash
deepseek-ai · deepseek reasoning efficient
#22 overall
- params
- 158B
- arch
- moe
- context
- 1024k
- license
- mit open weights
- released
- Apr 2026
- reasoning
- yes
- train compute
- 2.5e+24 FLOP
- downloads/30d
- 936.1k
# architecture
- layers
- 43
- d_model
- 4096
- heads
- 64q / 1kv
- head dim
- 512
- experts
- top-6 of 256 + 1 shared
- vocab
- 129280
- family
- deepseek_v4
signal path of one layer, generated from the registry's structured fields — dimension lines quote the model's real numbers; an MoE trace forks at the router, a dense trace runs straight through. Geometry fields come from the repo's config.json.
# why ranked
Overall: #22 — score 77.3 ● high — all signals present
| signal | weight | input (0–100) |
|---|---|---|
| aa_intelligence | 0.70 | 75.6 |
| bench_composite | 0.30 | 81.4 |
benchmark panel evidence:
| complete panel | score (0–100) |
|---|---|
| aa-coding-v1 | 85.2 |
| livebench-v1 | 77.6 |
Coding: provisional — score 75.0 ● low — a single signal; treat with caution
provisional: too few independent signals for a numbered position — the score is shown, but this model sorts after every ranked model.
| signal | weight | input (0–100) |
|---|---|---|
| aa_coding | 0.30 | 75.0 |
| aider_polyglot | 0.30 | — |
| swe_bench_verified | 0.40 | — |
missing signals are dropped and the remaining weights renormalized — never imputed.
# trend
methodology changed during this history window; score movement across that boundary is not model movement. See methodology v6.
# benchmarks
| benchmark | score | source | date |
|---|---|---|---|
| AA Coding Index | 56.2 | Artificial Analysis | — |
| AA Intelligence Index | 24.4 | Artificial Analysis | — |
| LiveBench Agentic Coding | 37.6 | LiveBench | Jun 2026 |
| LiveBench Coding | 69.2 | LiveBench | Jun 2026 |
| LiveBench Data Analysis | 68.0 | LiveBench | Jun 2026 |
| LiveBench IF | 63.1 | LiveBench | Jun 2026 |
| LiveBench Language | 70.1 | LiveBench | Jun 2026 |
| LiveBench Mathematics | 79.6 | LiveBench | Jun 2026 |
| LiveBench Reasoning | 70.6 | LiveBench | Jun 2026 |
# where to run
| provider | quant | ctx | $/M in | $/M out | $/M cache | price src | tps | uptime | ✓ |
|---|---|---|---|---|---|---|---|---|---|
| Featherless | — | — | — | — | via hfrouter | — | — | ||
| StreamLake | fp8 | 1000k | $0.08 | $0.17 | $0.02 | via openrouter | — | 100.0% | |
| DeepInfra | fp8 | 1024k | $0.09 | $0.18 | — | deepinfra | — | 100.0% | ✓ |
| GMICloud | fp8 | 1048k | $0.09 | $0.18 | $0.02 | via openrouter | — | 100.0% | |
| DigitalOcean | unknown | 1024k | $0.10 | $0.20 | $0.02 | via openrouter | — | 100.0% | |
| Fireworks | unknown | 1024k | $0.10 | $0.21 | $0.02 | via openrouter | — | 60.3% | |
| Relace | fp4 | 1024k | $0.05 | $0.45 | $0.05 | via openrouter | — | 100.0% | |
| SiliconFlow | fp8 | 1024k | $0.13 | $0.28 | $0.03 | openrouter | — | 100.0% | ✓ |
| Venice | unknown | 1000k | $0.14 | $0.28 | — | venice | — | 100.0% | ✓ |
| AtlasCloud | fp4 | 1024k | $0.14 | $0.28 | — | atlascloud | — | 99.7% | ✓ |
| Baidu | fp8 | 1024k | $0.14 | $0.28 | $0.03 | via openrouter | — | 98.4% | |
| Crusoe | fp8 | 1024k | $0.14 | $0.28 | $0.03 | via openrouter | — | 100.0% | |
| Nebius | 1024k | $0.14 | $0.28 | — | via litellm | — | — | ||
| Parasail | fp8 | 1024k | $0.14 | $0.28 | $0.07 | via openrouter | — | 100.0% | |
| Tencent | 1000k | $0.14 | $0.28 | — | via litellm | — | — | ||
| Tensormesh | 32k | $0.14 | $0.28 | — | via litellm | — | — | ||
| Wandb | 1049k | $0.14 | $0.28 | — | via litellm | — | — | ||
| Aihubmix | 1000k | $0.14 | $0.28 | — | via litellm | — | — | ||
| Prism | 1000k | $0.17 | $0.21 | — | via litellm | — | — | ||
| Alibaba | fp8 | 1000k | $0.15 | $0.30 | $0.03 | via openrouter | — | 95.9% | |
| Inferenceapis | 1024k | $0.19 | $0.38 | — | via litellm | — | — | ||
| Qwen Ai Platform | 1000k | $0.20 | $0.40 | — | via litellm | — | — | ||
| Qwencloud | 1000k | $0.20 | $0.40 | — | via litellm | — | — | ||
| DeepSeek | 1024k | $0.15 | $0.60 | $0.00 | via modelsdev | — | — | ||
| Mancer 2 | fp8 | 1024k | $0.19 | $0.50 | — | via openrouter | — | 97.0% | |
| Azure | unknown | 1024k | $0.21 | $0.56 | $0.03 | via openrouter | — | 99.1% | |
| Wafer | mxfp4 | 1024k | $0.60 | $0.17 | $0.20 | via openrouter | — | 100.0% | |
| Libertai | 200k | $0.25 | $1.75 | — | via litellm | — | — | ||
| Cloudflare | unknown | 375k | $0.44 | $1.32 | $0.01 | via openrouter | — | 99.9% |
sorted by blended price ((3·input + output) / 4 per 1M) · ✓ = the provider's own catalog confirms the offer · "via …" prices are what the aggregator routing the offer charges, not the provider's own list price
source aliases
- aa
deepseek-v4-flash-0420,deepseek-v4-flash-0420-high- epoch
DeepSeek-V4-Flash- hf
deepseek-ai/DeepSeek-V4-Flash-DSpark- openrouter
deepseek/deepseek-v4-flash