Skip to content
RankShark

DeepSeek V4 Flash 0731

DeepSeek V4 Flash 0731 is a sparse mixture-of-experts model from DeepSeek, with 13B active parameters out of 284B total.

ModelOpen sourceby DeepSeeksince 2026
46.4

SharkScore

#80 in Models

Why it scores 46.4

DeepSeek V4 Flash 0731 has a SharkScore of 46.4/100, ranked #80 of 608 in Models (updated 2026-11-19).

Each part compares DeepSeek V4 Flash 0731 with other listings in Models. Evidence covers 87% of the score's weight.

Adoption
38 · 30%
User reviews
no data · 25%
Maintenance
32 · 20%
Trust & transparency
80 · 15%
Momentum
no data · 10%

Paying never changes this score. How ranking works · method v2026-11-18.2

RankShark review of DeepSeek V4 Flash 0731

Automated evaluation by RankShark · 18 Nov 2026 · method v1.0 · How we review

We were not paid for this review.

As of 2026-11-18, DeepSeek V4 Flash 0731 is a text model listed at $0.06 per million input tokens and $0.18 per million output tokens, with a 1,048,576-token context window. [E2; MEASURED]

DeepSeek V4 Flash 0731 is a text-in, text-out model. The vendor describes it as a sparse mixture-of-experts model with 13B active parameters out of 284B total, and publishes instructions for local inference. [E1, E2]

Best for

  • Developers building coding or agent workflows. · model card, OpenRouter
  • Users who need a long context window for text input. · OpenRouter
  • Teams seeking local inference options through vLLM or SGLang. · model card

Not for

  • Workflows that require image, audio, or video input; the listing specifies text input and text output. · OpenRouter
  • Users who require a Jinja-format chat template; the vendor says this release does not include one. · model card

Key features

  • Long context. The listing gives a context length of 1,048,576 tokens; RankShark measured the same value on 2026-11-18. · OpenRouter
  • Reasoning effort controls. The vendor says reasoning effort supports low, high, and max. The OpenRouter listing also includes none. · model card, OpenRouter
  • Agent and coding use. The vendor presents benchmark results for coding and agent tasks and says the release has enhanced agentic capabilities. · model card
  • Speculative decoding. The vendor says the model includes a DSpark speculative decoding module and gives setup instructions for vLLM and SGLang. · model card
  • Local inference. The vendor provides instructions for running the model locally and examples using vLLM and SGLang. · model card

Pros

  • The listing reports a context window of 1,048,576 tokens. · OpenRouter
  • The vendor publishes local inference instructions and examples for vLLM and SGLang. · model card
  • The vendor provides a dedicated encoding folder with scripts and test cases for OpenAI-compatible message formats. · model card

Cons

  • The listing specifies text-only input and output. · OpenRouter
  • The vendor says the release does not include a Jinja-format chat template. · model card
  • No free plan or free trial stated on the pages we checked.
  • Whether customer data is used for model training is not stated on the pages we checked.

Pricing

PlanPriceLimits
OpenRouter listed token pricing$0.06 per million input tokens; $0.18 per million output tokens; $0.015 per million cached input tokensUsage-based token pricing; no monthly plan or quota stated.

Checked 18 Nov 2026

Free plan
Not stated (checked 18 Nov 2026)
Free trial
Not stated (checked 18 Nov 2026)
Trains on customer data
Not stated (checked 18 Nov 2026)
Data retention
Not stated (checked 18 Nov 2026)
Data regions
Not stated (checked 18 Nov 2026)
Data processing agreement
Not stated (checked 18 Nov 2026)
AI models used
DeepSeek-V4-Flash-0731 · model card, OpenRouter
API
The OpenRouter listing includes chat completions, completions, messages, and responses endpoints. · OpenRouter
SDKs
Not stated (checked 18 Nov 2026)
Webhooks
Not stated (checked 18 Nov 2026)
Platforms
The vendor provides local inference instructions with vLLM and SGLang, and identifies Transformers as the repository library. · model card
Security claims
No security certification stated (checked 18 Nov 2026)

What we measured

Context window (OpenRouter listing) · now
1,048,576 tokens
Price per million tokens (OpenRouter listing) · now
$0.06 in, $0.18 out
Package downloads · 30 days
2,019,920
How we tested · sources
  • Hourly uptime and response-time checks of the product's URL.
  • Repository, release and download data from GitHub, npm, PyPI and Hugging Face where they exist.
  • We did not sign up, pay, or use the product as a customer.
  • 18 Nov 2026: Plans or prices changed on the pricing page.
  • 18 Nov 2026: Plans or prices changed on the pricing page.

Something wrong? Report an error with a public link that shows it; corrections re-run the review.

Top alternatives to DeepSeek V4 Flash 0731

See all
  • 67.1

    Qwen3.8 27B is an open-weight dense vision-language model from Qwen.

    ModelOpen source
  • 65.5

    Qwen3.6-35B-A3B is an open-weight multimodal model from Alibaba Cloud with 35 billion total parameters and 3 billion active parameters per t

    ModelOpen source
  • 64.7

    Qwen3.8 Flash is a multimodal reasoning model from Alibaba.

    ModelOpen source
  • 64.3

    Qwen3.6 27B is a dense 27-billion-parameter language model from the Qwen Team at Alibaba, released in April 2026.

    ModelOpen source