Ember-1: Pricing, Context and Capabilities
fireworks/ember-1Ember-1 is Fireworks' reasoning model for text and image inputs.
1.05M tokens of context — about 786,432 words, on a scale that ends at 2M.
What it can do
- Reasoning
- Vision
- Tool use
- Takes text
- Takes image
- Gives text
Released 24 Sept 2026 · checked 3 Oct 2026.
About it
Ember-1 is a specialized reasoning model from Fireworks Research. It accepts text and images, can use tools, and returns text. The model has a 1,048,576-token context window and can produce up to 943,718 tokens in one response. Pricing is $3 per million input tokens and $15 per million output tokens.
Review
In a line
Ember-1 is a specialized reasoning model from Fireworks Research, built on Kimi K3. It takes text or images as input and produces text.
What it's good at
This model is aimed at reasoning tasks. It supports tool use and can work with images as well as text. Those capabilities may help when a task needs step-by-step analysis or image understanding. You can compare it with other reasoning models and vision models.
Price in plain terms
Input costs $3 per million tokens. Output costs $15 per million tokens, so generated text costs more than text sent to the model.
Reading a page of about 500 words costs about 0.19 cents. A chat turn with about 1,000 input tokens and 500 output tokens costs about $0.01. A thousand such turns cost about $10.50.
Context and limits
Ember-1 can take up to 1,048,576 tokens in one context, meaning the material it considers for a request. It can produce up to 943,718 tokens in one answer.
Its output is text only, even though it accepts images as input.
Ways to use it
Use the model ID fireworks/ember-1 when a service offers Ember-1. Send text, an image, or both, and receive a text response. It also supports tool use.
Verdict
Ember-1 suits people who need a reasoning-focused model that can read images, use tools, and handle very large contexts. Its input price is moderate compared with its output price, so frequent long answers can cost more. Choose it when those capabilities matter; note that output tokens cost more than input tokens.
Verdict
Choose Ember-1 for analysis that may involve images, tools, or unusually long source material. Its large context and high output limit leave room for lengthy requests and replies. Budget carefully for generated text, which is charged at $15 per million tokens, five times the input rate. It is less suitable when you need audio, video, or other non-text results, since its responses are text only. The model ID is fireworks/ember-1.
- Handles text and image inputs
- Supports reasoning and tool use
- 1,048,576-token context window
- Up to 943,718 output tokens
- About $0.01 for a typical chat turn
- Output costs $15 per million tokens
- Responses are text only
- A thousand typical chat turns cost about $10.50
- Pick it ifReasoning tasks using text or images; Long documents that need a large context; Workflows that use tools
- Skip it ifYou need audio, video, or other non-text output
Pricing
Ember-1 charges $3 per million input tokens and $15 per million output tokens. A 500-word page costs about 0.19 cents to read; a typical chat turn costs about $0.01.
Questions
What is Ember-1?
Ember-1 is a specialized reasoning model from Fireworks Research, built on Kimi K3.
Can Ember-1 understand images?
Yes. It accepts images and text as inputs.
Does Ember-1 return images?
No. Its output is text.
How much context can Ember-1 handle?
Its context window is 1,048,576 tokens.
What is Ember-1's maximum answer length?
It can produce up to 943,718 tokens in one response.
What does Ember-1 cost?
Input costs $3 per million tokens and output costs $15 per million tokens.
How much would 1,000 typical chat turns cost?
About $10.50, using the stated example of 1,000 input tokens and 500 output tokens per turn.
What model ID should I look for?
The model ID is fireworks/ember-1.