Back to Models
Reflection AI
beam-1.0
Chat only

Beam 1.0 501B A23B

Beam 1.0 is Reflection's open-weight reasoning model for coding and long-horizon agentic workflows. It uses a sparse mixture-of-experts architecture with 501B total and 23B active parameters per token, and a...

10/16/2026
1,048,576 tokens
Input: $0.422/M
Output: $1.27/M

Specifications

Modalities

Input
text
Output
text

Supported Parameters

frequency_penalty
include_reasoning
logit_bias
logprobs
max_tokens
min_p
presence_penalty
reasoning
reasoning_effort
repetition_penalty
response_format
seed
stop
structured_outputs
temperature
tool_choice
tools
top_k
top_logprobs
top_p

Max Output Tokens

943,718

Reasoning Configuration

Default
Thinking on (effort: max)
Selectable effort levels
max
xhigh
high
medium
low
Turning thinking off
Not possible (always thinks)

How to control thinking, and what it costs

Data policy

Prompt retention
Unknown — we could not confirm
Training
Not used for training

"Unknown" does not mean "safe". It means we could not confirm it.

Whether a provider trains on prompts is a declared value from our terms with them. Retention is determined from the upstream listing for every host this model can reach. Neither is guessed.

Available in the web chat only

Beam 1.0 501B A23B can be used in FastMetal's web chat, billed per token from your prepaid balance like any other model. It is not available through the API or MCP: a request naming it is refused by the gateway (HTTP 403).

Open in chat

How Beam 1.0 501B A23B actually answers

Real responses to our standard prompts, recorded on FastMetal.

Count the number of 'r's in 'strawberry'

Count the number of 'r's in 'strawberry'. Explain your reasoning step by step.

The word “strawberry” is spelled: s - t - r - a - w - b - e - r - r - y The letter “r” appears in these positions: - s**r**awberry (1st r) - strawbe**r**ry (2nd r) - strawber**r**y (3rd r) Therefore, there are **3** r's in “strawberry.”
7.6s response278 output tokens

Debug This Error

I'm getting the following error in my Node.js application: TypeError: Cannot read properties of undefined (reading 'map') at UserList (/app/components/UserList.js:12:25) at renderWithHooks (/app/node_modules/rea…

The error means `users` is `undefined` when `UserList` renders, so calling `.map` on `undefined` throws. ### Common causes 1. Parent renders `<UserList />` without a `users` prop. 2. Parent passes `users={data}` where `data` is still `undefined` (e.g., fetch not resolved). 3. Destructured prop name mismatch. ### Fixes **1. Default the prop** ```js function UserList({ users = [] }) { return ( <ul> {users.map(user => <li key={user.id}>{user.name}</li>)} </ul> ); } ``` **2. Guard in JSX** ```js {Array.isArray(users) && users.map(user => <li key={user.id}>{user.name}</li>)} ``` **3. Optional chaining** ```js {users?.map(user => <li key={user.id}>{user.name}</li>)} ``` Note…
12.8s response414 output tokens

Code Review

Please review the following Python function and suggest improvements for readability, performance, and best practices: def get_data(url, retries=3): import requests import time for i in range(retries):…

Here are some key issues and improvements: ### Main issues in the original - **Imports inside the function** run on every call and hurt readability. - **Bare `except:`** catches `KeyboardInterrupt`, `SystemExit`, and programming errors. - **Retries every non-200 response**, including `404`/`400`, which usually should not be retried. - **`item["active"] == True`** can raise `KeyError` and is stricter than necessary unless you truly need exact boolean equality. - **Fixed sleep and unnecessary final sleep** before returning. - **Repeated `item["name"]` / `item["id"]`** lookups are slightly slower and less clean than a comprehension. - No handling/validation for malformed JSON or non-list respo…
16.8s response1927 output tokens

Compare these answers side by side with other models →

Frequently asked questions

How much does the Beam 1.0 501B A23B API cost?
On FastMetal, Beam 1.0 501B A23B is billed per token: $0.422 per 1M input tokens and $1.27 per 1M output tokens on a US-dollar account, before tax. Usage is drawn from a prepaid balance; there is no subscription or monthly fee.
What is the context window of Beam 1.0 501B A23B?
1,048,576 tokens, shared between the prompt and the response.
What do I need to try Beam 1.0 501B A23B?
Create an account and add credit; the model is then available both in the browser chat and over the API. There is no contract or minimum spend.

Try Beam 1.0 501B A23B right now

Beam 1.0 501B A23B is available on FastMetal through one API key. Start in the browser, or call it from the OpenAI SDK.