Beam 1.0 501B A23B vs Perceptron Mk1 compared — price per token, context window, multimodality, openness and which to choose. Can the open-source model replace the paid one? Full 2026 breakdown.
Perceptron Mk1 — full profile ›
Prices & specs refreshed from live data · olud.ai
Open-model prices = cheapest provider via OpenRouter; official maker rates may be higher.
| Spec | Beam 1.0 501B A23B | Perceptron Mk1 | Winner |
|---|---|---|---|
| Maker | Reflection | Perceptron | – |
| Type | Open-weight | Proprietary | REBeam 1.0 501B A23B |
| Context window | 1M tokens | 33K tokens | REBeam 1.0 501B A23B |
| Input price | $0.25/M · free self-host | $0.15/M | PEPerceptron Mk1 |
| Output price | $1/M · free self-host | $1.5/M | REBeam 1.0 501B A23B |
| Vision / multimodal | No | Yes | PEPerceptron Mk1 |
| Tool / function calling | Yes | No | REBeam 1.0 501B A23B |
| Self-hostable | Yes | No (API only) | REBeam 1.0 501B A23B |
| License | Open weights | Proprietary | REBeam 1.0 501B A23B |
Beam 1.0 501B A23B is ~1.5× cheaper than Perceptron Mk1 on output tokens ($1 vs $1.5 per M tokens).
| Capability | Beam 1.0 501B A23B | Perceptron Mk1 |
|---|---|---|
| Open weights (downloadable) | ✓ | ✗ |
| Self-hostable | ✓ | ✗ |
| Runs fully offline | ✓ | ✗ |
| Vision / multimodal | ✗ | ✓ |
| Tool / function calling | ✓ | ✗ |
| 1M+ context window | ✓ | ✗ |
Independent benchmark scores measured by Artificial Analysis. Higher is better (except latency).
Benchmark data by Artificial Analysis.
| Criterion | Beam 1.0 501B A23B | Perceptron Mk1 |
|---|---|---|
| Cost-efficiency | 4.5 | 4.5 |
| Context window | 5.0 | 2.5 |
| Openness | 5.0 | 1.5 |
| Self-hosting | 5.0 | 1.0 |
| Multimodality | 3.5 | 3.5 |
Scores come from live data — output price (cost), context length, open vs closed weights (openness & self-hosting) and vision/tool support (multimodality). Raw task quality isn't scored here; it depends on your benchmark — see the verdict.
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...
Perceptron Mk1 (Mark One) is Perceptron's highest-quality vision-language model for video and embodied reasoning.** It accepts image and video inputs paired with natural language queries, and produces detailed visual understanding...
Beam 1.0 501B A23B is open-weight and competitive on many tasks, but Perceptron Mk1 may still lead on the hardest reasoning and agentic work. The gap keeps narrowing — benchmark both on your actual use case before deciding.
Yes. Beam 1.0 501B A23B has open weights, so you can self-host it on your own GPUs or run it via a low-cost API. Perceptron Mk1 is API-only and cannot be self-hosted.
Beam 1.0 501B A23B costs $1/M output vs $1.5/M for Perceptron Mk1 — roughly 2x cheaper via API, and free if you self-host.
Choose Beam 1.0 501B A23B if you want to self-host, keep your data private and skip per-token fees — it's open-weight and runs on your own hardware. Choose Perceptron Mk1 if you want frontier capability through a managed API with zero infrastructure to run.
Browse the full open-source model leaderboard, thousands of tools and live benchmarks — all in one place.
Open the leaderboard →