DeepSeek V4 Flash 0423
DeepSeek · Open weights · Released Apr 24, 2026 · deepseek/deepseek-v4-flash
1.05M contextToolsReasoningStructured outputPrompt cachingOpen weights
Available from 17 providersUSD / MTOK · VERIFIED 12 HOURS AGO
EST. MONTHLY COSTclick a row for detail · click headers to sort
| PROVIDER | INPUT | OUTPUT | CACHE | CONTEXT | QUANT | TTFT | TPS | SMOKE | UPTIME |
|---|---|---|---|---|---|---|---|---|---|
| STStreamLake | $0.049 | $0.098 | $0.01 | 1.02M | fp8 | 2.2s | 47 tok/s | 4/5 | 99.8% |
| GMIGMICloud | $0.059 | $0.118 | $0.012 | 1.05M | fp8 | 2.5s | 105 tok/s | 4/5 | 100.0% |
| DIDigitalOcean | $0.068 | $0.168 | $0.017 | 1.05M | — | 1.3s | 24 tok/s | 5/5 | 99.9% |
| SRSail Research | $0.09 | $0.18 | $0.02 | 1.05M | fp4 | 0.87s | 40 tok/s | 5/5 | 99.9% |
 DeepInfra | $0.09 | $0.18 | $0.018 | 1.05M | fp8 | 1.9s | 87 tok/s | 5/5 | 99.8% |
 SiliconFlow | $0.13 | $0.28 | $0.028 | 1.05M | fp8 | 1.8s | 56 tok/s | — | 99.9% |
 Alibaba | $0.134 | $0.268 | $0.027 | 1M | fp8 | 0.88s | 127 tok/s | — | 99.6% |
 Venice | $0.138 | $0.275 | $0.028 | 1M | — | 2.2s | 120 tok/s | — | 100.0% |
 Novita | $0.14 | $0.28 | $0.028 | 1.05M | fp8 | 0.79s | 111 tok/s | — | 99.8% |
| ATAtlasCloud | $0.14 | $0.28 | $0.028 | 1.05M | fp4 | 0.78s | 122 tok/s | — | 100.0% |
| BABaidu | $0.14 | $0.28 | $0.028 | 1.05M | fp8 | 0.73s | 138 tok/s | 4/5 | 99.3% |
| CWCoreWeave | $0.14 | $0.28 | $0.07 | 1.05M | fp8 | 1s | 63 tok/s | — | 99.6% |
 Parasail | $0.14 | $0.28 | $0.07 | 1.05M | fp8 | 1.2s | 34 tok/s | — | 99.8% |
| M2Mancer 2 | $0.14 | $0.45 | $0.014 | 1.05M | fp8 | 1.9s | 42 tok/s | — | 99.0% |
| PHPhala | $0.20 | $0.40 | $0.07 | 1.05M | — | 8.6s | 4 tok/s | — | 95.9% |
 Azure | $0.21 | $0.56 | $0.031 | 1.05M | — | 18.9s | 51 tok/s | — | 98.2% |
 Cloudflare | $0.44 | $1.32 | $0.014 | 384K | — | 0.76s | 78 tok/s | — | 100.0% |
TTFT/TPS: median (P50) over trailing 72h · chat10k profile (~10K in / 200 out) · measured via OpenRouter routing · SMOKE: of 5 programmatic capability probes passed (JSON schema, instruction following, tool call, long-context, benign compliance; retry-once) · screened via OpenRouter routing · v0.1 · methodology v1.1 · measured 43 hours ago
Input price history · best listed price, observed
↓ 28% SINCE AUG 22, 2026
$0.068 · AUG 22, 2026$0.049 · TODAY
Pricing variants · USD / MTok in / out
STANDARD
$0.049 / $0.098
per MTok in / out
CACHE READ $0.01CACHE WRITE Free
Recent changesRSS ↗
Aug 24, 2026price decreased · input0.0503 → 0.0489
Aug 24, 2026price decreased · output0.1005 → 0.0977
Aug 23, 2026price decreased · input0.056 → 0.0503
Aug 23, 2026price decreased · output0.112 → 0.1005
Aug 23, 2026price decreased · input0.0601 → 0.056
Aug 23, 2026price decreased · output0.1201 → 0.112
Aug 22, 2026price decreased · input0.0679 → 0.0601
Aug 22, 2026price decreased · output0.1481 → 0.1201
About
DeepSeek V4 Flash is an efficiency-optimized Mixture-of-Experts model from DeepSeek with 284B total parameters and 13B activated parameters, supporting a 1M-token context window. It is designed for fast inference and...
Specifications
Context window1,048,576
Max output384,000
Modalities intext
Modalities outtext
LicenseOpen weights
ReleasedApr 24, 2026
TokenizerDeepSeek
WeightsHugging Face ↗
Related
Embed badge
[](https://modelindex.ai/models/deepseek/deepseek-v4-flash)Sources
OpenRouter API12 hours ago
AGGREGATORLiteLLM dataset12 hours ago
AGGREGATORHugging Face12 hours ago
COMMUNITYModelIndex benchmarks43 hours ago
MEASUREDVerified 12 hours ago · list prices, not negotiated rates.