DeepSeek logo

DeepSeek AI models

12 DeepSeek models on FastRouter, all behind one OpenAI-compatible API. Compare pricing, context windows and benchmarks, then open any model for its providers and code samples.

Filter in catalog
Models
12
Largest context
1.05M
DeepSeek V4.1 Flash Abliterated
Lowest input /1M
$0.06
V4 Flash
Top intelligence
32.6
DeepSeek V4.1 Flash 1017

All DeepSeek models

DeepSeek logo

Abliterated Large Flash is Abliteration's hosted version of DeepSeek-V4.1-Flash-1017 with its refusal behavior removed through abliteration. It accepts text and images (up to four images per request history, 12 MB each, in PNG, JPEG, WebP, or GIF) and returns text, with a 1M-token context window and up to 262,144 output tokens including reasoning. Reasoning can be turned off or set to low, high, or max. It supports tool calling, JSON mode and JSON Schema output, streaming, and web search, through OpenAI Chat Completions, OpenAI Responses, and Anthropic Messages request shapes. It is US-hosted with zero prompt retention. Video input is rejected.

deepseek/deepseek-v4.1-flash-abliteratedOct 23, 2026
Context
1.05M
Price /1M
$1.00 in$2.00 out
Intel
—
DeepSeek logo

DeepSeek-V4.1-Flash-1017 is the October 17, 2026 update to DeepSeek-V4.1-Flash, a multimodal Mixture-of-Experts model that natively accepts text and images, with a 1M-token context window and MIT license. It improves reasoning, coding, agentic tasks, and visual benchmarks over V4.1 Flash while reducing task execution time and output length, so tasks finish faster and cheaper. DeepInfra serves it in fp8 with JSON output, function calling, and zero data retention, plus Priority and Flex service tiers. DeepSeek reports DeepSWE v1.1 at 77.0 (up from 74.2) and Terminal-Bench 4.0 at 42.3 (up from 31.2) versus V4.1 Flash.

deepseek/deepseek-v4.1-flash-1017Oct 17, 20262.7s latency
Context
1.05M
From /1M
$0.20$0.60
$0.10 in$0.29 out
Intel
32.6
DeepSeek logo

DeepSeek-V4.1-Flash is a multimodal Mixture-of-Experts model with a 552B-parameter backbone, activating just 8B parameters per token during prefill and 16B during decoding. It natively processes images and text, supports a 1M-token context window, and introduces a Causal Encoder-Decoder architecture with Compressed Sparse Attention 2 to make long-context inference more efficient. Configurable reasoning-effort levels let developers trade latency and cost for deeper deliberation. Compared with DeepSeek-V4-Flash-0731, it reduces the global KV-cache footprint by roughly 4x while delivering stronger coding, reasoning, and agentic performance.

deepseek/deepseek-v4.1-flashSep 14, 202611s latency
Context
1.05M
From /1M
$0.20$0.60
$0.14 in$0.42 out
Intel
—
DeepSeek logo

DeepSeek V4.1 Flash Ultrafast is OpenGateway's high-throughput serving tier of DeepSeek's open-weight V4.1 Flash multimodal reasoning model. It accepts text and images and returns text, with a 1M-token context window, configurable reasoning effort, and tool calling through an OpenAI-compatible chat completions endpoint. Pricing is tiered, with higher rates once total input exceeds 256K tokens.

deepseek/deepseek-v4.1-flash-ultrafastSep 10, 2026
Context
1.05M
Price /1M
$0.32 in$1.30 out
Intel
—
DeepSeek logo

DeepSeek-V4-Flash-Vision-Exp is an experimental multimodal model in the V4-Flash family with a 1M-token context, image input with visual grounding, tool calling, JSON output and configurable reasoning effort.

deepseek/deepseek-v4-flash-vision-expAug 21, 20269.6s latency
Context
1.05M
From /1M
$0.44$1.32
$0.22 in$0.65 out
Intel
—
DeepSeek logo

DeepSeek V4 Pro 0813 is the general-availability release of DeepSeek V4 Pro, a 1.6T-parameter Mixture-of-Experts model (49B active) with a 1M-token context window, hybrid reasoning and tool calling.

deepseek/deepseek-v4-pro-0813Aug 12, 2026
Context
1.05M
Price /1M
$1.32 in$3.96 out
Intel
23.1
DeepSeek logo

DeepSeek V4 Flash is an efficiency-optimized Mixture-of-Experts large language model from DeepSeek with a 1M-token context window and strong general, coding, and reasoning performance. It is designed as a fast, low-cost frontier model suitable for long-context chat, agents, and tools. The Flash variant emphasizes high throughput and low latency while maintaining competitive reasoning quality.

deepseek/deepseek-v4-flashJul 31, 2026
Context
1.05M
Price /1M
$0.06 in$0.18 out
Intel
—
DeepSeek logo

DeepSeek-V4-Pro is a 1.6 trillion parameter Mixture-of-Experts (MoE) language model from DeepSeek AI, with 49 billion parameters activated per token and support for a 1 million token context window. It features a hybrid attention architecture combining Compressed Sparse Attention (CSA) and Heavily Compressed Attention (HCA), which reduces inference FLOPs to 27% and KV cache to 10% of DeepSeek-V3.2 at 1M context, enabling 83.5% MRCR comprehension. The model was pre-trained on over 32 trillion tokens using the Muon optimizer and a two-stage post-training process involving SFT, RL with GRPO, and on-policy distillation for domain-specific expertise.

deepseek/deepseek-v4-proApr 24, 202620s latency
Context
1.05M
Price /1M
$1.30 in$2.60 out
Intel
23.1
DeepSeek logo

DeepSeek-V3.2 is positioned as a next-generation “general-purpose + reasoning” model intended to be a daily driver at roughly frontier (GPT‑5-class) performance on broad tasks like coding, math, agents, and general chat. It is released in several variants (such as V3.2, V3.2-Exp, and V3.2-Speciale), sharing the same core architecture but targeting different trade‑offs between efficiency and maximum reasoning power.

deepseek/deepseek-v3.2Dec 1, 202513s latency
Context
164K
Price /1M
$0.26 in$0.38 out
Intel
—
DeepSeek logo

DeepSeek V3.1 is an advanced open-source hybrid AI model designed to balance powerful reasoning with high-speed efficiency. It uniquely supports two inference modes—"Think" for deep reasoning and "Non-Think" for direct, lightweight tasks—making it versatile across use cases. Built with a mixture-of-experts architecture, it scales to 685B parameters while activating only 37B per token, enabling cost-effective performance. With a 128K context window, it can handle large documents, codebases, and complex multi-step workflows. DeepSeek V3.1 is optimized for tool use, agent-based applications, and enterprise deployment through open weights and developer-friendly APIs.

deepseek/deepseek-v3.1Aug 21, 20252.3s latency
Context
164K
Price /1M
$0.25 in$0.95 out
Intel
—
DeepSeek logo

DeepSeek-R1-0528 is a minor version upgrade of the DeepSeek R1 reasoning model.

deepseek/deepseek-r1-0528May 28, 2025
Context
164K
Price /1M
$0.50 in$2.15 out
Intel
—
DeepSeek logo

DeepSeek-V3 is a strong Mixture-of-Experts language model with 671B total parameters (37B active per token), using Multi-head Latent Attention and DeepSeekMoE for efficient inference.

deepseek/deepseek-v3Dec 26, 202425s latency
Context
164K
Price /1M
$0.32 in$0.89 out
Intel
—

Frequently asked questions

Model scores sourced from ArtificialAnalysis

DeepSeek AI Models: API Pricing & Benchmarks | FastRouter.ai