- Size
- 501B parameters, 23.0B active (from the repo name)counted from the safetensors header on the Hub
- Licence
- unknown, Apache License 2.0Our reading, check the licence: You can use, change and redistribute the model, including commercially. Keep the licence and any notices, and say which files you changed; you do not have to publish your changes. It includes a patent licence that ends for anyone who sues over patents in the work.
- On the Hub since
- 7 Oct 2026the date the repo was created
- Context length
- 1M tokensfrom the model's config
- Kind
- Base model (no base model declared on the Hub)BeamForCausalLM
- On Hugging Face
- reflection/Beam-1.0-501B-A23B13k downloads in the last 30 days, 325 likes
Scores
No independent scores for this model in Epoch AI's data. A score is only shown here when Epoch's name for a model matches this repo with certainty; an uncertain match is left out rather than guessed.
Claimed by the maker Claimed
The model card has no claimed scores in a machine-readable form (the card's model-index), so none are shown. The maker may give scores in the card text: read the card.
Other views, which we link to and do not copy: look this model up on Artificial Analysis and the Arena leaderboard.
Run it yourself
Memory needed = weights + context cache + 0.50 GB for the runtime. Context cache at 8k tokens: 0.37 GB, worked out from the model's config (reflection/Beam-1.0-501B-A23B), at 16-bit precision. Choose a machine above to see a verdict for each format. Method
GGUF for llama.cpp, Ollama, LM Studio and others
bartowski/Beam-1.0-501B-A23B-GGUF (converted by bartowski), 5.6k downloads. Sizes are the real files on the Hub, all parts added up.
| Format | Size | Memory needed | Get it |
|---|---|---|---|
| IQ3_XXS | 196.6 GBfile size, 6 parts | 197.5 GB196.6 + 0.37 + 0.50 | Commandsllama.cpp llama-server -hf bartowski/Beam-1.0-501B-A23B-GGUF:IQ3_XXSDownload hf download bartowski/Beam-1.0-501B-A23B-GGUF --include "Beam-1.0-501B-A23B-IQ3_XXS/Beam-1.0-501B-A23B-IQ3_XXS-*-of-00006.gguf" --local-dir Beam-1.0-501B-A23B-GGUF |
| IQ3_XS | 209.1 GBfile size, 6 parts | 210.0 GB209.1 + 0.37 + 0.50 | Commandsllama.cpp llama-server -hf bartowski/Beam-1.0-501B-A23B-GGUF:IQ3_XSDownload hf download bartowski/Beam-1.0-501B-A23B-GGUF --include "Beam-1.0-501B-A23B-IQ3_XS/Beam-1.0-501B-A23B-IQ3_XS-*-of-00006.gguf" --local-dir Beam-1.0-501B-A23B-GGUF |
| Q3_K_S | 210.4 GBfile size, 6 parts | 211.2 GB210.4 + 0.37 + 0.50 | Commandsllama.cpp llama-server -hf bartowski/Beam-1.0-501B-A23B-GGUF:Q3_K_SDownload hf download bartowski/Beam-1.0-501B-A23B-GGUF --include "Beam-1.0-501B-A23B-Q3_K_S/Beam-1.0-501B-A23B-Q3_K_S-*-of-00006.gguf" --local-dir Beam-1.0-501B-A23B-GGUF |
| Q3_K_M | 220.4 GBfile size, 7 parts | 221.2 GB220.4 + 0.37 + 0.50 | Commandsllama.cpp llama-server -hf bartowski/Beam-1.0-501B-A23B-GGUF:Q3_K_MDownload hf download bartowski/Beam-1.0-501B-A23B-GGUF --include "Beam-1.0-501B-A23B-Q3_K_M/Beam-1.0-501B-A23B-Q3_K_M-*-of-00007.gguf" --local-dir Beam-1.0-501B-A23B-GGUF |
| Q3_K_L | 232.4 GBfile size, 7 parts | 233.3 GB232.4 + 0.37 + 0.50 | Commandsllama.cpp llama-server -hf bartowski/Beam-1.0-501B-A23B-GGUF:Q3_K_LDownload hf download bartowski/Beam-1.0-501B-A23B-GGUF --include "Beam-1.0-501B-A23B-Q3_K_L/Beam-1.0-501B-A23B-Q3_K_L-*-of-00007.gguf" --local-dir Beam-1.0-501B-A23B-GGUF |
| IQ3_M | 238.1 GBfile size, 7 parts | 239.0 GB238.1 + 0.37 + 0.50 | Commandsllama.cpp llama-server -hf bartowski/Beam-1.0-501B-A23B-GGUF:IQ3_MDownload hf download bartowski/Beam-1.0-501B-A23B-GGUF --include "Beam-1.0-501B-A23B-IQ3_M/Beam-1.0-501B-A23B-IQ3_M-*-of-00007.gguf" --local-dir Beam-1.0-501B-A23B-GGUF |
| IQ4_XS | 254.3 GBfile size, 8 parts | 255.2 GB254.3 + 0.37 + 0.50 | Commandsllama.cpp llama-server -hf bartowski/Beam-1.0-501B-A23B-GGUF:IQ4_XSDownload hf download bartowski/Beam-1.0-501B-A23B-GGUF --include "Beam-1.0-501B-A23B-IQ4_XS/Beam-1.0-501B-A23B-IQ4_XS-*-of-00008.gguf" --local-dir Beam-1.0-501B-A23B-GGUF |
| Q4_0 | 264.6 GBfile size, 8 parts | 265.5 GB264.6 + 0.37 + 0.50 | Commandsllama.cpp llama-server -hf bartowski/Beam-1.0-501B-A23B-GGUF:Q4_0Download hf download bartowski/Beam-1.0-501B-A23B-GGUF --include "Beam-1.0-501B-A23B-Q4_0/Beam-1.0-501B-A23B-Q4_0-*-of-00008.gguf" --local-dir Beam-1.0-501B-A23B-GGUF |
| Q4_K_S | 273.2 GBfile size, 8 parts | 274.1 GB273.2 + 0.37 + 0.50 | Commandsllama.cpp llama-server -hf bartowski/Beam-1.0-501B-A23B-GGUF:Q4_K_SDownload hf download bartowski/Beam-1.0-501B-A23B-GGUF --include "Beam-1.0-501B-A23B-Q4_K_S/Beam-1.0-501B-A23B-Q4_K_S-*-of-00008.gguf" --local-dir Beam-1.0-501B-A23B-GGUF |
| IQ4_NL | 290.5 GBfile size, 9 parts | 291.4 GB290.5 + 0.37 + 0.50 | Commandsllama.cpp llama-server -hf bartowski/Beam-1.0-501B-A23B-GGUF:IQ4_NLDownload hf download bartowski/Beam-1.0-501B-A23B-GGUF --include "Beam-1.0-501B-A23B-IQ4_NL/Beam-1.0-501B-A23B-IQ4_NL-*-of-00009.gguf" --local-dir Beam-1.0-501B-A23B-GGUF |
| Q4_1 | 292.5 GBfile size, 9 parts | 293.4 GB292.5 + 0.37 + 0.50 | Commandsllama.cpp llama-server -hf bartowski/Beam-1.0-501B-A23B-GGUF:Q4_1Download hf download bartowski/Beam-1.0-501B-A23B-GGUF --include "Beam-1.0-501B-A23B-Q4_1/Beam-1.0-501B-A23B-Q4_1-*-of-00009.gguf" --local-dir Beam-1.0-501B-A23B-GGUF |
| Q4_K_M | 292.9 GBfile size, 9 parts | 293.8 GB292.9 + 0.37 + 0.50 | Commandsllama.cpp llama-server -hf bartowski/Beam-1.0-501B-A23B-GGUF:Q4_K_MDownload hf download bartowski/Beam-1.0-501B-A23B-GGUF --include "Beam-1.0-501B-A23B-Q4_K_M/Beam-1.0-501B-A23B-Q4_K_M-*-of-00009.gguf" --local-dir Beam-1.0-501B-A23B-GGUF |
| Q5_K_S | 330.7 GBfile size, 9 parts | 331.6 GB330.7 + 0.37 + 0.50 | Commandsllama.cpp llama-server -hf bartowski/Beam-1.0-501B-A23B-GGUF:Q5_K_SDownload hf download bartowski/Beam-1.0-501B-A23B-GGUF --include "Beam-1.0-501B-A23B-Q5_K_S/Beam-1.0-501B-A23B-Q5_K_S-*-of-00009.gguf" --local-dir Beam-1.0-501B-A23B-GGUF |
| Q5_K_M | 354.6 GBfile size, 10 parts | 355.5 GB354.6 + 0.37 + 0.50 | Commandsllama.cpp llama-server -hf bartowski/Beam-1.0-501B-A23B-GGUF:Q5_K_MDownload hf download bartowski/Beam-1.0-501B-A23B-GGUF --include "Beam-1.0-501B-A23B-Q5_K_M/Beam-1.0-501B-A23B-Q5_K_M-*-of-00010.gguf" --local-dir Beam-1.0-501B-A23B-GGUF |
| Q6_K_S | 391.8 GBfile size, 11 parts | 392.6 GB391.8 + 0.37 + 0.50 | Commandsllama.cpp llama-server -hf bartowski/Beam-1.0-501B-A23B-GGUF:Q6_K_SDownload hf download bartowski/Beam-1.0-501B-A23B-GGUF --include "Beam-1.0-501B-A23B-Q6_K_S/Beam-1.0-501B-A23B-Q6_K_S-*-of-00011.gguf" --local-dir Beam-1.0-501B-A23B-GGUF |
| Q6_K | 411.4 GBfile size, 12 parts | 412.3 GB411.4 + 0.37 + 0.50 | Commandsllama.cpp llama-server -hf bartowski/Beam-1.0-501B-A23B-GGUF:Q6_KDownload hf download bartowski/Beam-1.0-501B-A23B-GGUF --include "Beam-1.0-501B-A23B-Q6_K/Beam-1.0-501B-A23B-Q6_K-*-of-00012.gguf" --local-dir Beam-1.0-501B-A23B-GGUF |
| Q8_0 | 496.2 GBfile size, 14 parts | 497.1 GB496.2 + 0.37 + 0.50 | Commandsllama.cpp llama-server -hf bartowski/Beam-1.0-501B-A23B-GGUF:Q8_0Download hf download bartowski/Beam-1.0-501B-A23B-GGUF --include "Beam-1.0-501B-A23B-Q8_0/Beam-1.0-501B-A23B-Q8_0-*-of-00014.gguf" --local-dir Beam-1.0-501B-A23B-GGUF |
MLX for Apple silicon
| Format | Size | Memory needed | Repo |
|---|---|---|---|
| MLX nvfp4OpensourceWTF/Beam-1.0-501B-A23B-NVFP4-q8-mlx-stream | 12.4 GBfile size | 13.2 GB12.4 + 0.37 + 0.50 | Download page |
Original weights size from the safetensors header, for vLLM, SGLang, Transformers
| Format | Size | Memory needed | Repo |
|---|---|---|---|
| BF16original weights | 933.8 GBfile size | 934.7 GB933.8 + 0.37 + 0.50 |
hf download reflection/Beam-1.0-501B-A23B --local-dir Beam-1.0-501B-A23BOther conversions (7; sizes not read)
- reflection/Beam-1.0-501B-A23B-NVFP4 NVFP4, by the maker, 5.9k downloads
- reflection/Beam-1.0-501B-A23B-FP8 FP8, by the maker, 4.0k downloads
- Baekpica/Beam-1.0-501B-A23B-VQ-GGUF GGUF, 6.3k downloads; sizes not used: every file is far too small for this model; the repo may be tagged wrongly or hold only part of it
- Bespoke-Intelligence/Beam-1.0-501B-A23B-NVFP4-TP3-3x-DGX-Spark NVFP4, 518 downloads
- zouaynib/Beam-1.0-501B-A23B-MXFP4 Other, mxfp4, 336 downloads
- bertholomus/Beam-1.0-501B-A23B-EXL3-3bpw-2xGB10 EXL3, 308 downloads
- diffbot/Beam-1.0-501B-A23B-EXL3-3bpw-2x-RTX-PRO-6000 EXL3, 160 downloads
The Hub tags 12 quantised repos as conversions of this model; we read the most downloaded and show the maker's own first.
Run it by API
5 providers serve this model through the Hugging Face router. Prices are US dollars per million tokens. Precision is what the provider reports; a lower precision than the original weights can change answers. Live first, then cheapest output.
| Provider | In | Out | Context | Precision | Tokens/s | First token | Uptime 24h | Tools | JSON |
|---|---|---|---|---|---|---|---|---|---|
| CoreWeave | $0.30 | $1.00 | 1M | nvfp4 | 202 | 0.5 s | 97.0% | yes | yes |
| Nebius | $0.25 | $1.00 | 1M | fp4 | 384 | 0.5 s | 98.8% | yes | yes |
| BaseTen | $0.40 | $1.20 | 1M | fp4 | 528 | 0.2 s | 99.0% | no | yes |
| DigitalOcean | $0.40 | $1.20 | 1M | not given | 226 | 0.3 s | 99.8% | yes | yes |
| Together | $0.40 | $1.20 | 1M | not given | 345 | 0.2 s | 97.4% | yes | yes |
Through the router, with a Hugging Face token (the router bills your Hugging Face account). Change the name after the colon to pick another provider.
curl https://router.huggingface.co/v1/chat/completions \
-H "Authorization: Bearer $HF_TOKEN" -H "Content-Type: application/json" \
-d '{"model": "reflection/Beam-1.0-501B-A23B:coreweave", "messages": [{"role": "user", "content": "Hello"}]}'Not built yet: one key for every open model, with BlockRun, and our own tested-quality score per provider. Until then, precision and measured speed are the best guide to a provider here.
Planned for this page Not built yet
- Try it in private chat
- A link to Solana AI's confidential chat for the models it serves. This model is not listed as served there.
- Fine-tune it
- Fine-tuning a few base models from this page and publishing the result here.
- Creators and earnings
- Pages for people who publish fine-tunes and adapters of this model, and what they earn.
- Fund and bounties
- Bounties for new open models, paid from a public fund. How the fund will work.
- Contribute compute
- How to give compute to open model work, with Akash.
Model data from the Hugging Face Hub 19 Nov 2026, 13:43 UTC. The repo. Found a mistake on this page? Tell us.