Ember-1

What is Ember-1?

Ember-1 is a specialized reasoning model from Fireworks Research, built on Kimi K3.

Per 1M input / output tokens

$3.00 / $15.00

Starting price

Context window

1.05M

Up to 944K output per reply

Input

text + image

Output: text

What Ember-1 can do

• Multi-step reasoning

• Tool-using agents

• Structured outputs (JSON)

• Image understanding

Ember-1 vs similar models: price

What you pay for Ember-1 and the models most like it: the same kind of work, from the families we feature, priced within about three times of it.

• Ember-1 vs Gemini 3.1 Pro Preview

  Gemini 3.1 Pro Preview costs $6.30: 30% less than Ember-1.

  −10% at the top tier

  Gemini 3.1 Pro Preview API 

• Ember-1 vs Claude Opus 5.5

  Claude Opus 5.5 costs $10.80: 1.2× the price of Ember-1.

  −10% at the top tier

  Claude Opus 5.5 API 

• Ember-1 vs GPT-6.1 Sol

  GPT-6.1 Sol costs $5.40: 40% less than Ember-1.

  −10% at the top tier

  GPT-6.1 Sol API 

• Ember-1 vs Qwen3.8 Max

  Qwen3.8 Max costs $3.80: 58% less than Ember-1.

  −25% at the top tier

  Qwen3.8 Max API 

• Ember-1 vs GLM 5.3

  GLM 5.3 costs $2.465: 73% less than Ember-1.

  −25% at the top tier

  GLM 5.3 API 

Ember-1 specs

The technical card for Ember-1, from the OceanToken catalog.

Context window

1.05M tokens

1,048,576

Max output

944K tokens

943,718

Input

text, image

Output

text

Reasoning

Yes

reasoning_effort

Tool calling

Yes

tool_choice

Structured outputs

Yes

response_format

Prompt caching

Reads $0.30

official price, at every tier

Ember-1 model ID and endpoint

OceanToken model ID

fireworks/ember-1Copied

OpenAI-compatible base URL

https://api.oceantoken.ai/v1Copied

Endpoint

POST /v1/chat/completionsCopied

Ember-1 API pricing

Prices come from the OceanToken catalog as of November 5, 2026. You pay per use from a prepaid balance, with no subscription.

Ember-1 cost calculator

Pick a typical request or enter your own token counts. The calculator uses the starting prices on this page and shows the same workload on the models compared above.

At the starting price (tier 1). Cached input is billed at the cache-read price; cache writes are left out because they depend on how often your prompt prefix changes. Your console shows the rate for your tier.

What a request costs

Input and output tokens at the starting price above, with no caching.

How to use Ember-1 on OceanToken

Three steps from no account to a working request. The same key and balance work for every model in the catalog.

1. 1

   Create an account and get an API key

   Create an OceanToken account, then create a key. One key works for every model your account can use.

   Get API Key 

2. 2

   Add credits

   OceanToken is prepaid: add credits, and each request is billed from that balance. Your console shows the rate for your tier.

   Add credits 

3. 3

   Call Ember-1 from your code

   Point an OpenAI SDK at https://api.oceantoken.ai/v1 and use the model ID fireworks/ember-1. The code below sends one request.

   from openai import OpenAI
   # Your API Key from the OceanToken console
   client = OpenAI(base_url="https://api.oceantoken.ai/v1",
                   api_key="YOUR_API_KEY")
   reply = client.chat.completions.create(
       model="fireworks/ember-1",
       messages=[{"role": "user", "content": "Hello"}])
   print(reply.choices[0].message.content)

   Specs, request sample and alternatives in the docs