# DeepSeek V4 is live on Fireworks - Explore All Models Available via API on Fireworks

> Fireworks supports Deepseek models for both inference & fine-tuning, delivering blazing-fast speeds and up to 40% cost reduction from closed-source models. Deepseek, a pioneer in open-weights reasoning, offers a focused family of LLMs, including the MoE-based DeepSeek-V3 and reasoning-first Deepseek-R1. Fireworks supports Deepseek models for both inference and fine-tuning, delivering blazing-fast speeds (up to 273 tokens/sec) with proprietary Fire Attention CUDA kernels and up to 40% cost reduction when migrating from expensive closed-source models. Leverage day-zero model availability and an OpenAI-compatible API to accelerate development for use cases like code development and AI Agents with tool use.

## DeepSeek 4 is Live on Fireworks

Founded in 2023, DeepSeek is the primary pioneer of the open-weights reasoning movement, committed to democratizing advanced AI through radical efficiency. On April 24, DeepSeek released its latest flagship model, DeepSeek V4. Developers can now start building with this frontier open-source model with a single click today.

## DeepSeek Model Collection

Models from DeepSeek that harmonize high computational efficiency with superior reasoning and agent performance. 

[Explore all DeepSeek models in the model library](https://fireworks.ai/models?providers=Deepseek)

- **[DeepSeek V4.1 Flash 1017](https://fireworks.ai/models/deepseek-ai/deepseek-v4p1-flash-1017)**: Deepseek · $0.3/M Input · $1.2/M Output · 1,048,576 Context · Vision
- **[DeepSeek V4.1 Flash](https://fireworks.ai/models/deepseek-ai/deepseek-v4p1-flash)**: Deepseek · $0.3/M Input · $1.2/M Output · 1,048,576 Context · Vision
- **[DeepSeek-V4-Flash-Vision-Exp](https://fireworks.ai/models/deepseek-ai/deepseek-v4-flash-vision-exp)**: Deepseek · 1,048,576 Context · Vision
- **[DeepSeek-V4-Pro-0813](https://fireworks.ai/models/deepseek-ai/deepseek-v4-pro-0813)**: Deepseek · 1,048,576 Context · LLM
- **[DeepSeek-V4-Flash-0731](https://fireworks.ai/models/deepseek-ai/deepseek-v4-flash-0731)**: Deepseek · 1,048,576 Context · LLM
- **[DeepSeek R1 0528 Distill Qwen3 8B](https://fireworks.ai/models/fireworks/deepseek-r1-0528-distill-qwen3-8b)**: Deepseek · 131,072 Context · LLM
- **[DeepSeek Prover V2](https://fireworks.ai/models/fireworks/deepseek-prover-v2)**: Deepseek · 163,840 Context · LLM
- **[DeepSeek Coder V2 Instruct](https://fireworks.ai/models/fireworks/deepseek-coder-v2-instruct)**: Fireworks · 32,768 Context · LLM

## DeepSeek Innovations That Break the Mold

- **[The DeepSeek-R1 Breakthrough](https://fireworks.ai/blog/deepseek-r1-deepdive)**: First open-source model to rival OpenAI o1 in complex reasoning, math, and coding. Production-ready and battle-tested.
- **Architectural Gamechangers**: Mixture-of-Experts (MoE) plus Multi-head Latent Attention (MLA) deliver higher throughput, lower cost, and superior multi-step reasoning. · [Learn more about MoE](https://fireworks.ai/blog/deepseek-model-architecture#more-aggressive-mixture-of-expert)
- **Reasoning You Can Rely On**: Reinforcement learning ensures stable and predictable outcomes for coding, agent workflows, and planning.
- **[Research Driven, Production Ready](https://www.linkedin.com/feed/update/urn:li:activity:7412696675913428992/)**: Innovations like mHC are baked directly into shipped models. Always ahead, always scalable.

## Why choose DeepSeek on Fireworks?

### Unmatched Speed That Powers Your Workflows

Consistently rank first in TTFT and TPS benchmarks, achieving up to **389 tokens per second** with our proprietary performance optimizations.

### Slash Costs Without Sacrificing Quality

Up to 40% Savings: Migrate from expensive closed-source models while keeping output quality high.

### Day Zero Deployments Gets You to Market Faster

- Instant Availability: Start testing and deploying new DeepSeek models within hours of launch
- Effortless Migration: Switch quickly with our OpenAI-compatible API, often just a two-line code change,

## Sourcegraph cuts latency 30% and boosts acceptance rates 2.5X

CASE STUDY

Sourcegraph accelerated bug resolution and improved code quality at scale by applying context-aware, streaming AI edits tailored to their codebase. The ability to quickly evaluate DeepSeek models sped up iteration cycles and avoided vendor lock-in.

[Read the case study](https://fireworks.ai/blog/story-sourcegraph-code-generation)

**2\.5X** Higher Fix Acceptance · Sourcegraph

## Get Started

- **Quick Deployment**: Deploy in minutes with a single click on our serverless platform. · [Start now](https://docs.fireworks.ai/getting-started/quickstart)
- **Tailored AI**: Own your model's performance and behavior by getting started with Fine Tuning for your specific use case. · [Learn more about fine tuning](https://docs.fireworks.ai/fine-tuning/finetuning-intro#supported-models)
- **Get Expert Support**: Contact our team to discuss your workload, and let us help you optimize for performance and cost. · [Contact us](https://fireworks.ai/contact)
