Experiential
ModelsObservability
Star us on GitHubJoin the DiscordDocs
Sign in
Continue with GoogleContinue with GitHub
or
Models

DeepSeek V4 Flash 0731

deepseek-v4-flash-0731by DeepSeekZDR enforceable
Try in Chat

DeepSeek V4 Flash 0731 is a sparse mixture-of-experts model from DeepSeek, with 13B active parameters out of 284B total. This re-post-trained revision is suited for coding, reasoning, and agent workflows....

Context

1.05M

Max output

393K

Price per million tokens

$0.44in$0.028cached$1.32out

Released

Jul 2026

Waterfall

  1. 1T
    Tencent Cloud via Experiential Cloud$0.44/$1.3299.9% up--
  2. 2N
    Novita via Experiential Cloud$0.44/$1.3299.9% up--
  3. 3
    Qwen Cloud via Experiential Cloud$0.44/$1.3299.9% up--
  4. 4W
    Wafer via Experiential Cloud$0.44/$1.3299.9% up--
  5. 5
    reka via Experiential Cloud$0.090/$0.2599.9% up--
  6. 6
    abstract via Experiential Cloud$0.10/$0.5599.9% up--

Quickstart

I want you to route my LLM calls for "deepseek-v4-flash-0731" through the Experiential gateway instead of calling the provider directly. It speaks the OpenAI Chat Completions API, so this is a base-URL and key swap. Please:

1. Point the client at https://api.experientiallabs.jp/v1 as the base URL.
2. Securely ask me for my API key. Authenticate with my Experiential key from EXPLABS_API_KEY. I can create one under Settings -> API Keys.
3. Use the model id "deepseek-v4-flash-0731" exactly.
4. Update every place my code builds an LLM client for this model to use that base URL and key, leaving streaming and tool-calls as they are.
5. Make one test call and show me the reply plus the token usage, so we confirm it runs on my Experiential credits.

Tell me which files you changed.
CapabilitiesAccepts text

Sampling

Temperature*0-1.99Top-p*0.01-1Top-k*Stop sequences*

Tools & structure

Tools*Strict tools*Structured output*

Streaming

Streaming

Limits

Max-tokens fieldmax_tokens

Sending an unsupported field? See error reference

Benchmarks5
BenchmarkScoreSource
GPQA Diamondpublic leaderboard · Aug 202689.9%public leaderboard · Aug 2026
SWE-bench Verifiedpublic leaderboard · Aug 202688.8%public leaderboard · Aug 2026
LiveBenchpublic leaderboard · Aug 202674.2%public leaderboard · Aug 2026
Humanity's Last ExamHugging Face · Aug 202637.8%Hugging Face · Aug 2026
Terminal-BenchHugging Face · Aug 202682.7Hugging Face · Aug 2026
Providers serving this model8
Status
abstract via Experiential Cloudexperientialno datano data99.9%#2 / 61.05M393K$0.10$0.55$0.10-no dataactiveestimated
reka via Experiential Cloudexperientialno datano data99.9%#1 / 61.05M393K$0.090$0.25cache$0.036-no dataactiveestimated
NNovita via Experiential Cloudexperientialno datano data99.9%#4 / 61.05M393K$0.44$1.32cache$0.028-ZDRactive-
Qwen Cloud via Experiential Cloudexperientialno datano data99.9%#5 / 61.05M393K$0.44$1.32cache$0.028-no dataactive-
TTencent Cloud via Experiential Cloudexperientialno datano data99.9%#3 / 61.05M393K$0.44$1.32cache$0.028-no dataactive-
WWafer via Experiential Cloudexperientialno datano data99.9%#6 / 61.05M393K$0.44$1.32cache$0.028-no dataactive-
Azure FoundryFW-DeepSeek-V4-Flash-0731no datano data99.9%-1.05M393K$0.14≈$0.28≈cache$0.028≈-no dataactiveestimated
Azure FoundryDeepSeek-V4-Flash-0731-2026-07-31deployed 2026-07-31no datano data99.9%-1.05M393K$0.14≈$0.28≈cache$0.028≈-no dataactiveestimated
Hugging Face