Sulat.com
AI models
$10 off the fastest DeepSeek V4.1 Flash, Kimi K3 and GLM 5.3 from Synthetic
Kilo Gateway logo

Model details

DeepSeek: DeepSeek V4 Flash 0731 (free)

Within the DeepSeek model family identified as deepseek-flash, the "V4 Flash 0731 (free)" entry appears in third-party catalogs such as Infron AI and Pluely Pro, which aggregate hundreds of large language models from providers including OpenAI, Google, Anthropic, Meta, xAI, and DeepSeek for browsing and selection. The "(free)" tag and the date-based 0731 suffix suggest a time-stamped free-tier snapshot intended for cost-free experimentation rather than a paid production deployment, and its placement in marketplaces that bundle over 280 models indicates it is positioned as an accessible, general-purpose text generation option alongside reasoning-oriented alternatives.

Infron AI's marketplace surfaces DeepSeek-V4-0731 alongside a separate "limited-time 70% off offer" promotion, signaling that the broader V4 family has both free and discounted commercial variants available through resellers. Because Kilo, the listing gateway here, advertises access to "500+ models, zero markup" via its open-source coding agent and gateway, models routed through it are intended for developer-facing workflows, including coding assistants such as Claude Code, Cline, and Codex integrations that Infron highlights as supported clients for comparable entries. Readers should treat the Flash 0731 free variant as a market-listed text model whose qualitative behavior is best judged from direct use rather than from third-party catalogs alone.

Kilo Gatewaydeepseek/deepseek-v4-flash-0731:freedeepseek-flash

Quick Info

Powered by
Provider
Kilo Gateway
Model key
deepseek/deepseek-v4-flash-0731:free
Release date
Jul 31, 2026
Last updated
Jul 31, 2026
Knowledge cutoff
2025-05
Input modalities
Output modalities
Capabilities

Cost

A provider subscription or plan supersedes token-based pricing for this model.

Limits

Output tokens
393,216 tokens
Context window
1,048,576 tokens

Latest news about DeepSeek: DeepSeek V4 Flash 0731 (free)

Kilo Gateway

CoverageBenchmark

DeepSeek V4 Flash 0731 (free) is the GA release of DeepSeek's V4 Flash, a sparse mixture-of-experts model that activates 13B parameters from a 284B-parameter total, with a 1.0M-token context window and a stated focus on coding, reasoning, and agent workflows. Per the OpenRouter model page excerpt, DeepSeek — the model Beyond the core model attributes, the excerpt also reports provider-routed benchmark scores (GPQA Diamond roughly 89–91% and TAU-Bench up to 77.5% across listed providers), a 1-week average throughput of about 34 tok/s at P50, and an availability figure of 99.64% over the September 15–18 window. These operational numbe

Videos about DeepSeek: DeepSeek V4 Flash 0731 (free)

More models around DeepSeek: DeepSeek V4 Flash 0731 (free)