All models
Reflection
Beam 1.0 501B A23B fastruns locally
reflection/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...
Input / 1M
$0.25
USD per million tokens
Output / 1M
$1.00
USD per million tokens
Est. cost / token
$0.000000437
$0.44 per 1M, 3:1 mix
Context window
1.05M
max output 131K
Major features
Open weightsCachingReasoningJSON schemaTools
Task suitability (heuristic 1–5, not a benchmark)
- coding3
- reasoning3
- math3
- summarization4
- extraction4
- classification4
- chat3
- creative3
- translation4
- vision1
- long context5
- agentic3
Run it locally
Will it run on my machine? →| Quantization | Quality | Memory needed (8K ctx) |
|---|---|---|
| Q2_K | poor | 227 GB |
| Q3_K_M | fair | 264 GB |
| Q4_K_Mrecommended | good | 328 GB |
| Q5_K_M | very good | 381 GB |
| Q6_K | excellent | 438 GB |
| Q8_0 | near-lossless | 566 GB |
| F16 | lossless | 1059 GB |
Estimates. Mixture-of-experts: memory covers all parameters; only the active ones are used per token, so it runs faster than its size suggests.
Price history
| Recorded | Input | Output | Cached | Source |
|---|---|---|---|---|
| 2026-11-05 03:20 | $0.25 | $1.00 | $0.050 | openrouter |
A row is added whenever a sync detects a price change. Vendor pricing source
Pricing (USD per 1M tokens)
- Input
- $0.25
- Output
- $1.00
- Cached input
- $0.050
- Blended (3:1)
- $0.44
Details
- Company
- Reflection
- Version
- 1
- Family
- Beam
- Released
- 2026-10-16
- Parameters
- 501.3B (23B active)
- Max output
- 131K
- Open weights
- Yes
- Inputs
- text
- Last synced
- 46 min ago