Fireworks Ember-1 cuts Kimi K3 token use by ~40%
Token waste in agentic loops is one of the biggest hidden costs of running reasoning models, and post-training on real task loops offers a cheaper fix than prompt tricks or output caps. Teams building coding agents on K3 or similar models can cut inference spend materially without dropping success rates. Fireworks is US-based but the base model is Moonshot (Kimi)'s K3, a useful data point in the open-model cost race.
Fireworks Research released Ember-1, a post-trained version of Kimi K3 focused on reducing redundant reasoning.
The model uses roughly 40% fewer tokens than K3 while matching its benchmark performance.In a live A/B test on coding traffic, Ember-1 used 71% fewer reasoning tokens and 39% fewer total tokens at the same success rate.
Fireworks RL trained the model on real agentic coding task loops to teach it which reasoning changes the answer versus which is repetition.
Reasoning models can spend over 90% of output tokens on thinking before answering, which is costly in agentic loops.
Ember-1 by the Fireworks Research team is built on Kimi K3, and uses ~40% fewer tokens while achieving the same performance on benchmarks.
This was accomplished by post-training K3 to think less repetitively.
Reasoning models spend most of their output tokens (sometimes 90%+) on thinking before they answer, which gets expensive in agentic loops where the model tends to re-think the same thoughts on every step.
Fireworks RL trained on real agentic coding task loops to teach the model which reasoning actually changes the answer vs which is just looping.
In a live A/B test on coding traffic, Ember-1 used 71% fewer reasoning tokens and 39% fewer total tokens than K3 at the same success rate.
This report is categorized under AITOP's "Models" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to US-China foundation model benchmarks and commercialization comparison and AI developer toolchains, coding agents, and engineering workflows. Key entities and technical topics under continuous tracking include Ember-1, Kimi K3, Fireworks, reasoning models.
Source attribution: curated from Cline (@cline). Cross-source corroboration: 6 independent sources have covered this development (including elvis, IT之家, Sebastian Raschka), confirming industry-wide signal strength.
