Ember-1 is a new specialized model from Fireworks Research that delivers Kimi K3’s quality with 40% fewer tokens. Built on Kimi K3, it learned to cut unnecessary reasoning while keeping the thinking that matters. We tested it on external benchmarks, in live customer A/B tests, and on our own coding and agent workloads, and quality held up in every setting. Available today, Ember-1 kicks off an ongoing series of specialized models by Fireworks, shaped by what developers want next.
How Fireworks Research built Ember-1
We heard from users that they needed K3’s coding capabilities at a lower cost, because its long reasoning traces made automated coding expensive at scale. Turning down K3’s reasoning effort didn’t solve this. Lower effort settings gave up too much quality. To keep the quality and cut the tokens, the model had to learn to reason more efficiently, and that meant training it.
Getting there took serious research. Our team ran more than 50 training experiments and over 200 evaluations, and developed new training algorithms along the way to shorten reasoning without losing accuracy. We did it all on Fireworks Serverless Training.
We trained across a broad set of tasks so the token savings would carry over to many workloads. We used our own data, and no customer data to train this model. We then evaluated Ember-1 on the Specialized Intelligence Index, public benchmarks, and live production traffic to confirm it used fewer tokens with no drop in quality.
The problem: thinking models think too much
Reasoning models like Kimi K3 spend the majority of their generated tokens, sometimes more than 90%, on internal reasoning rather than the answer itself. This thinking structure is expensive on a single request, but it gets much worse in multi-turn agentic workloads. Every turn replays all prior reasoning back to the model, so context grows roughly quadratically with the number of turns.
Our experiments said the reasoning Kimi K3 emits is far longer than the task requires, and the excess can be removed without touching the answer.
Benchmark results
On Doximity’s Bedside Bench, a physician-validated benchmark spanning 500 clinical cases across 10 specialized categories, Ember-1 set a new Pareto frontier across both open and closed models including GPT-5.6 Sol, GPT-6 Astra, and Claude Opus 5 on cost/task.
| Industry Benchmarks | N | K3 max | Ember-1 | Token reduction |
|---|---|---|---|---|
| Terminal Bench 2.1 | 89 | 80.9% | 82.0% | -51.9% |
| SWE-bench Verified | 500 | 93.2% | 92.2% | -15.5% |
| SWE-Interact | 75 | 21.3% | 20.0% | -32.5% |
| DeepSWE 1.1 | 113 | 66.4% | 75.2% | -23.7% |
Customer validation
We ran live A/B tests with two customers on their production coding workloads. In both cases, Ember-1 delivered approximately 35% fewer tokens per task at comparable quality. Following the A/B tests, one customer is now running Ember-1 in live production, with plans to scale it up to replace the base model entirely.
| Score | Steps | Output Tokens | Total token reduction | |
|---|---|---|---|---|
| Kimi K3 | 0.751 | 23.8 | 49.3K | - |
| Ember-1 | 0.753 | 21.4 | 29.9K | 39% |
Ember-1 is rolling out as a serving option alongside the base Kimi K3 model as a Research Preview release on Serverless.