# Fireworks adds DeepSeek V4.1 Flash 1017 to its API

> The model is available through Fireworks’ serverless API. The company describes it as faster and stronger, but gives no benchmark figures on the listing.

Oossa · 2026-10-23 · https://oossa.com/en/fireworks-adds-deepseek-v4-1-flash-1017-to-its-api

Fireworks AI lists DeepSeek-V4.1-Flash-1017 as a serverless API model, with a context window of 1,048,576 tokens. The listing calls it a minor update and says it improves reasoning, coding, agent tasks and visual benchmarks while reducing execution time and output length.

Fireworks charges $0.30 per million input tokens and $1.20 per million output tokens; cached input tokens cost $0.006 per million. The listing does not provide benchmark scores or details of who ran the evaluations.

## The facts

- Context window: 1,048,576 tokens.
- Serverless pricing: $0.30 per million input tokens, $0.006 per million cached input tokens and $1.20 per million output tokens.

## Why it matters

Developers using Fireworks can try the model through its serverless API at published per-token rates, but the listing’s improvement claims cannot be assessed from benchmark results because it provides no scores or evaluation details.

## Sources & references

1. [DeepSeek V4.1 Flash 1017NewVision • Serverless$0.3/M Input • $1.2/M Output • 1048576 Context](https://fireworks.ai/models/deepseek-ai/deepseek-v4p1-flash-1017/) – Fireworks AI
2. [Deepseek's latest AI model is smarter than ever, but its low running costs come at a price](https://www.tomshardware.com/tech-industry/artificial-intelligence/deepseeks-latest-ai-model-is-smarter-than-ever-but-its-low-running-costs-come-at-a-price-open-weight-model-is-fit-for-agentic-workloads-but-is-more-prone-to-hallucination-than-proprietary-competitors) – Tom's Hardware, 2026-10-23

Last updated: 2026-10-23