Fireworks · Blogs

Blogs listings from Fireworks on TechiSeek. 148 public listings.

• Fireworks RFT: Build AI agents with fine-tuned open models that outperform frontier closed models (Fireworks)
  Fireworks RFT (Reinforcement Fine-Tuning) is a managed service that enables developers, enterprises, and researchers to fine-tune open mode…
• Trilogy’s Playbook for Open-Weight Cybersecurity with Kimi K3 (Fireworks)
  Trilogy’s new AI Cybersecurity Playbook centers on Kimi K3, leveraging open weights to ensure defenders have reliable, high-volume inferenc…
• Optimizing Llama 4 Maverick on Fireworks (Fireworks)
  Minutes after Meta published the weights, the model showed up in the Fireworks catalogue (accounts/fireworks/models/llama4-maverick-instruc…
• How Harvey & Fireworks Beat Closed Source on Cost + Quality (Fireworks)
  We explore two system-level techniques on Harvey’s Legal Agent Benchmark that reduce reliance on single frontier model calls while reaching…
• Inside Fireworks Multi-region Deployments: One Deployment, Global Scale (Fireworks)
  Use state-of-the-art, open-source LLMs and image models at blazing fast speed, or fine-tune and deploy your own at no additional cost with…
• Building AI agents with the Fireworks Experimentation Platform (GA) and Build SDK (Beta) (Fireworks)
  To access the full capabilities of the experimentation platform, we’re also excited to announce the beta of our brand-new Build SDK, a tool…
• Deployment Shapes: One Click Deployment Configured for You (Fireworks)
  Deployment Shapes offer tailored LLM inference infrastructure, allowing developers to optimize for latency, throughput, or cost.
• NVIDIA Nemotron 3 Nano on Fireworks: The Engine for Next-Generation AI Agents (Fireworks)
  The model builds on the Nemotron 2 Nano release combining a new Mixture-of-Experts architecture with the Nemotron hybrid transformer-mamba…
• Agentic AI Systems (Fireworks)
  AI is evolving from passive responders into proactive agents that can perceive, reason, and act autonomously.
• Fine-Tune Your Own Embedding Model from an LLM (Fireworks)
  Use state-of-the-art, open-source LLMs and image models at blazing fast speed, or fine-tune and deploy your own at no additional cost with…
• Fireworks and AMD Partner to Power Next-Generation AI Infrastructure on AMD Instinct™ GPUs (Fireworks)
  Fireworks and AMD have entered a multi-year partnership to optimize AMD Instinct™ GPUs and accelerate AI adoption across enterprises and de…
• Fireworks is moving to prepaid billing on July 1st (Fireworks)
  Use state-of-the-art, open-source LLMs and image models at blazing fast speed, or fine-tune and deploy your own at no additional cost with…
• Frontier RL Is Cheaper Than You Think (Fireworks)
  Cross-region RL changes the economics of frontier training: compact checkpoint deltas, hot-load updates, and bounded staleness let rollout…
• Fireworks Dev Day 2025 Wrapped (Fireworks)
  Yesterday (May 28th), we hosted our very first Fireworks DevDay, and what an incredible day it turned out to be.
• The Missing Piece of the OpenClaw Mania: Truly ‘Own Your AI’ with Fireworks (Fireworks)
  Use state-of-the-art, open-source LLMs and image models at blazing fast speed, or fine-tune and deploy your own at no additional cost with…
• Fireworks Achieves Triple ISO Certification, Giving Enterprises Full Control and Trust in AI (Fireworks)
  Use state-of-the-art, open-source LLMs and image models at blazing fast speed, or fine-tune and deploy your own at no additional cost with…
• Introducing FLUX.1 Kontext on Fireworks (Fireworks)
  Learn about the FLUX.1 Kontext in-context image generation models. See examples of their editing, character consistency, and style transfer…
• Fireworks Expands AWS Alliance: Strategic Collaboration Agreement + GenAI Competency (Fireworks)
  Fireworks has signed a Strategic Collaboration Agreement (SCA) with AWS and achieved the AWS Generative AI Competency Partner status.
• Introducing Llama 3.1 inference endpoints in partnership with Meta (Fireworks)
  Introducing Llama 3.1 405B, one of the largest open-source models, featuring a 128K context length, support for 10 languages, and powerful…
• Distillation with Reasoning: Can DeepSeek R1 Teach Better Than Humans? (Fireworks)
  Chain-of-Thought (CoT) reasoning improves model output quality but increases costs. Through distillation, this reasoning ability can be tra…
• How Fireworks evaluates quantization precisely and interpretably (Fireworks)
  With the release of Llama 3.1, there’s been considerable discussion about the benefits and tradeoff of different quantization methods.
• Firefunction-v2: Function calling capability on par with GPT4o at 2.5x the speed and 10% of the (Fireworks)
  Use state-of-the-art, open-source LLMs and image models at blazing fast speed, or fine-tune and deploy your own at no additional cost with…
• DeepSeek V4 Pro: Tops SWE-Bench & Cuts Cost per Task by 3x vs. Fable 5 (Fireworks)
  Use state-of-the-art, open-source LLMs and image models at blazing fast speed, or fine-tune and deploy your own at no additional cost with…
• Building Enterprise-Scale RAG Systems with Fireworks and MongoDB Atlas (Fireworks)
  In the fast-paced world of enterprise data, extracting actionable insights from vast amounts of unstructured information is a challenge man…
• Fireworks.ai Now Available on LangChain Prompt Playground (Fireworks)
  With the popularity of large language models (LLMs) such as ChatGPT and Llama 2, there are now a multitude of models to choose from — inclu…
• Qwen3 Decoded: Choosing the Right Model For Your Task (Fireworks)
  Compare Qwen3 Instruct, Thinking, and Coder. Learn which models are best suited for various tasks based on architecture, features, and benc…
• Unlock Advanced Reasoning with NVIDIA Nemotron Nano 2 Models on Fireworks (Fireworks)
  NVIDIA Nemotron Nano 2 on Fireworks: Expert Reasoning, Unprecedented Efficiency Announcing the launch of NVIDIA Nemotron Nano 2 9B and 12B…
• How we fixed prompt injection for all models on Fireworks (Fireworks)
  The blog highlights a critical prompt injection vulnerability in LLM serving stacks where user input containing model control tokens can ov…
• Fireworks f1: A breakthrough in complex reasoning with Compound AI (Fireworks)
  At Fireworks, we believe the future of AI is shifting to compound AI systems that combine specialized models and tools to achieve better pe…
• Trilogy Validates Open-Weight AI Models for Enterprise Workloads with Fireworks (Fireworks)
  Trilogy’s AI Center of Excellence evaluated open-weight models to address rising inference costs, rate limits, and operational constraints…
• FireFunction V1 - Fireworks’ GPT-4-level function calling model - 4x faster than GPT-4 and open (Fireworks)
  Last December, Fireworks launched the alpha version of our first function calling model to empower developers to incorporate external knowl…
• Muse Glimmer from Meta on Fireworks: Ideal for your Always-On Agents (Fireworks)
  Meta’s Muse Glimmer 30B model is now available on Fireworks, delivering high-concurrency, cost-effective performance for always-on AI agent…
• 50 Trillion Tokens Per Day: The State of Agent Environments (Fireworks)
  Agents and LLMs have scaled massively, processing 1.5 quadrillion tokens per month. The focus is shifting from model intelligence to the co…
• Building an open-source Browser Agent on Fireworks (Fireworks)
  In this technical deep dive, we'll explore how large language models (LLMs) can be given the ability to "see" web content and take actions…
• How Gumloop Scaled Open-Weight Model Usage 7x in 3 Weeks with Fireworks (Fireworks)
  Gumloop scaled its open-weight model usage 7x in three weeks by partnering with Fireworks.
• From text to task: Constrained generation for structured extraction in R1 (Fireworks)
  Constrained generation ensures AI stays on track, delivering clear, structured, and reliable responses.
• Announcing custom models and on-demand H100s with 50%+ lower costs and latency than vLLM (Fireworks)
  At Fireworks, we’re empowering developers to productionize generative AI with unparalleled speed, quality and cost.
• Reinforcement Fine Tuning: Train expert open models to surpass closed frontier models (Fireworks)
  We’re excited to announce the beta release of Reinforcement Fine-Tuning (RFT), a powerful new technique to create expert models for complex…
• FireAttention V3: Enabling AMD as a viable alternative for GPU inference (Fireworks)
  This time we are going to focus on a different GPU hardware, namely AMD MI300 GPU. While spec-wise it looks quite superior to NVIDIA H100 G…
• New in Fireworks: Image-to-Image and ControlNet support for SSD-1B and SDXL! (Fireworks)
  The Fireworks.ai blazing-fast inference platform enables developers to build with generative AI to accelerate product innovation.
• DeepSeek R1 Just Got Eyes with Fireworks Document Inlining (Fireworks)
  A smart reasoning LLM is good, but a smart reasoning VLM is better! We’re excited to demonstrate how DeepSeek R1, a state-of-the-art reason…
• MiniMax M3 is live: long context + native multimodality at 1/20th the price (Fireworks)
  MiniMax M3 is now available on Fireworks with Day-0 support — delivering the first open-weight model to combine frontier-level coding, nati…
• Eval Protocol: RL on your agents, in any environment (Fireworks)
  Eval Protocol (EP) is an open-source, language-agnostic framework for reinforcement fine-tuning (RL) on agents in production environments,…
• Quality first: how Fireworks.ai is the go-to place for gpt-oss (Fireworks)
  Use state-of-the-art, open-source LLMs and image models at blazing fast speed, or fine-tune and deploy your own at no additional cost with…
• Deep-Dive into LLM Fine-Tuning (Fireworks)
  LLM fine-tuning improves accuracy, speed, and cost efficiency for specialized use cases. Get in-depth guidance for optimal fine-tuning with…
• Gen-1 Slides: Opus 5-level decks at a fraction of the cost (Fireworks)
  Genspark partnered with Fireworks Lab to post-train MiniMax M3 into Gen-1 Slides, a model that produces Opus 5-level decks at roughly one-t…
• Three Tests to Run Before You Switch from LoRA to FullFT (Fireworks)
  When Full Fine-Tuning beats LoRA, the adapter may not be the problem. Controlled experiments on Qwen3.5-9B show how data coverage, optimiza…
• Mistral Small 3 Now Available on Fireworks: Faster, Lighter, and More Efficient (Fireworks)
  The latest open-weight model from Mistral—Mistral Small 3—is now live on Fireworks! Fireworks is excited to partner with Mistral to be an o…
• Reinforcement learning: Why alignment of numerics and MoE routing matter (Fireworks)
  Frontier training requires co-optimized training and inference: Fireworks develops and validates the trainer and rollout engine together so…
• Optimizing Retrieval Augmented Generation (RAG) with MongoDB Atlas and Fireworks (Fireworks)
  In this blog, we precisely help you kickstart your Generative AI Application development Journey and how to build a Retrieval Augmented Gen…