Ember 1
Ember-1 is a specialized reasoning model from Fireworks Research, built on [Kimi K3](https://openrouter.ai/moonshotai/kimi-k3).
ember-1Released 24 Sep 2026 (8 weeks ago)Pricing
List prices at Fireworks AI, in US dollars per million tokens.
| Tier | Input / 1M | Output / 1M | |
|---|---|---|---|
| Standard | $3.00 | $15.00 | |
| Cached input 90% off input | $0.30 | 90% off input | |
| Priority 25% more | $3.75 | $18.75 | 25% more |
What it costs a month
At Fireworks AI list prices, 30 days.
- Chat assistant5,000 requests a day, 2K in / 0.5K out, cached$1,782
- RAG over documents2,000 requests a day, 8K in / 0.6K out, cached$1,332
- Coding agent1,000 requests a day, 40K in / 2K out, cached$2,232
- Overnight batch20,000 requests a day, 4K in / 0.4K out$10,800
Where to run it
1 host serve Ember 1. The comparison is on blended price, 3 parts input to 1 part output.
| Host | In / out per 1M | Model id | |
|---|---|---|---|
Fireworks AIInference ember-1 | $3.00 / $15.00 | ember-1 |
Capabilities
Core
- Streaming
- Function calling
- JSON mode
- Structured outputs
Multimodal
- Vision
- Image input
- Audio
- Image generation
- Video
Advanced
- Reasoning
- Tool calling
- MCP
- Prompt caching
- Batch API
- Embeddings
- Takes in
- TextImages
- Produces
- Text
- Features
- Prompt caching
Cheaper alternatives
Current models that cost less and keep most of what this one can do: tools, vision, reasoning, structured output, caching and enough context.
Same model, cheaper host
No host sells it for less than Fireworks AI.
Smaller Independent models
No cheaper Independent model keeps its key capabilities.
Other frontier labs
- 20% less
- 33% less
- 33% less
- 50% less
Open-weight picks
- 33% less
- 64% less
- 71% less
- 91% less
Call it
Using Fireworks AI's API.
from litellm import completion
response = completion(
model="fireworks_ai/ember-1",
messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)Quick answers
How much does Ember 1 cost?
On Fireworks AI, Ember 1 costs $3.00 per million input tokens and $15.00 per million output tokens. Cached input costs $0.30 per million tokens.
Which APIs offer Ember 1?
Ember 1 runs on 1 API: Fireworks AI.
What is Ember 1's context window?
1M tokens.
Is Ember 1 being retired?
No retirement date is scheduled for it on any API we track.
What are cheaper alternatives to Ember 1?
Models that cost less and keep most of what it can do: GPT-5.3 Codex on OpenAI (20% less), Gemini 4 Carbon on Google AI Studio (33% less) and Kimi K3.1 Pro on Databricks (33% less). Savings are on a mix of 3 input tokens to 1 output token.
When was Ember 1 released?
Independent released it on 24 Sep 2026.
What is Ember 1's API model id?
Use ember-1 with Fireworks AI.
Sources
Raw catalog entry for ember-1 on Fireworks AI
{
"cache_read_input_token_cost": 3e-7,
"cache_read_input_token_cost_priority": 3.75e-7,
"input_cost_per_token": 0.000003,
"input_cost_per_token_priority": 0.00000375,
"litellm_provider": "fireworks_ai",
"max_input_tokens": 1048576,
"mode": "chat",
"output_cost_per_token": 0.000015,
"output_cost_per_token_priority": 0.00001875,
"source": "https://api.fireworks.ai/v1/serverless/models?format=nested",
"supports_function_calling": true,
"supports_mid_conversation_system": true,
"supports_prompt_caching": true,
"supports_reasoning": true,
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true
}