Fireworks AI retires dataset creation, upload, and download APIs
Fireworks AI retired its training dataset creation, upload, and download APIs, returning HTTP 501 for dataset storage actions while keeping list, get, and delete active.
Also shipped Nov 15, 2026Fireworks AI in November 2026 →
Looker Gemini-assisted features will be billed starting February 1, 2027
Usage of Gemini-assisted features in Looker that exceeds instance token allocations will be subject to billing starting February 1, 2027. This applies to capabilities like Conversational Analytics.
PostgreSQL integrated CDC pipeline is now in Beta
Databricks now offers an integrated CDC pipeline for PostgreSQL, allowing data ingestion without a separate gateway. This feature is currently in Beta.
Improved comments in the preview and on shared preview links
Comments in the preview have several updates: Each thread has a number, such as #3 , so you can refer to it when you discuss feedback. To attach a thread to a different element, drag its pin onto that element. To see what is under an open thread, drag the thread aside. You can read and reply to threads in the comments sidebar instead of next to their pins.
Qwen Code Release v0.27.4
Qwen Code v0.27.4 enables Managed Agent sessions to access granted local directories and open interactive Web Terminals without persisting credentials.
Qwen Code SDK TypeScript Release v0.1.34
Alibaba released Qwen Code SDK TypeScript version 0.1.34, which bundles Qwen Code CLI version 0.27.4 built from source.
Qwen Code Desktop v0.27.4
Alibaba released Qwen Code Desktop v0.27.4, adding client filesystem registration into managed sessions, Feishu custom bot webhooks, and security fixes.
More from Fireworks AI platformEvery Fireworks AI platform release →
Fireworks AI releases Training Cookbook on PyPI and migrates repository
Fireworks AI published fireworks-training-cookbook 2026.11.5.1 on PyPI for Python 3.11+ and migrated its training cookbook to fw-ai-external/cookbook.
Inference New Serverless mode: Ultra-fast (Preview) Ultra-fast is a new high-speed Serverless mode, now in...
Inference New Serverless mode: Ultra-fast (Preview) Ultra-fast is a new high-speed Serverless mode, now in Preview, for highly latency-sensitive applications. Ultra-fast variants aim for 6-8x generation speed, and up to 10x for agentic coding tasks. Like Fast, it is not a different model, so output quality is the same.
Training Serverless Training deprecation: DeepSeek V4 Flash (0731) DeepSeek V4 Flash (0731) is removed from...
Training Serverless Training deprecation: DeepSeek V4 Flash (0731) DeepSeek V4 Flash (0731) is removed from the shared Serverless Training pool effective November 2, 2026. Migrate Serverless Training workloads to DeepSeek V4.1 Flash 1017 . This change applies only to Serverless Training.
Fireworks AI adds Rune 26B-A4B model to Serverless
Fireworks AI has added the Rune 26B-A4B model to its Serverless platform. This decision model processes documents and questions to return calibrated probabilities without generating text, and is billed based on input tokens only.
Errors identify their source
Fireworks AI updated its error reporting to provide more granular source identification for failures, including new error codes for parsing and transport issues.
Unparseable trajectories are discarded
Fireworks AI now discards unparseable trajectories to prevent them from impacting training data quality.
Failure traces are saved locally
Failed model trajectories now generate private JSONL traces in the system temporary directory. These traces include request data, token IDs, and exception chains to help developers identify whether failures originated from the model, tokenizer, or renderer.
Image inputs for FiretitanSamplingClient.sample
The sample and sample_async methods now support a keyword-only images argument. This allows developers to pass base64 data URLs for token prompts using placeholders.