LocalFit
Submit a benchmark

Comet_60B_A12B_BF16

MultiverseComputingCAI/Comet_60B_A12B_BF16

ReleasedRelease date unknownSize61B MoE · 12B activeRoleAssistantLicensenvidia-nemotron-open-model-licenseContext1M config max

Where it runs 8K context · ≥ 5 t/s floor

SetupBest planMemoryGenerate
8 GB GPU + 32 GB RAMDoes not fit
12 GB GPU + 32 GB RAMDoes not fit
16 GB GPU + 64 GB RAMDoes not fit
24 GB GPU + 64 GB RAMDoes not fit
32 GB GPU + 128 GB RAMnative F32/BF16vLLMGPU streams weights from RAM · 84 GiB in RAMFits*≤ 1.7 t/s Ceiling
2× 24 GB GPUs + 128 GB RAMnative F32/BF16vLLMGPU streams weights from RAM · 92 GiB in RAMFits*≤ 1.6 t/s Ceiling
96 GB workstation GPU + 256 GB RAMnative F32/BF16vLLMGPU streams weights from RAM · 20 GiB in RAMFits*≤ 6.9 t/s Ceiling
Mac, 24 GBDoes not fit
Mac, 48 GBDoes not fit
Mac, 128 GBDoes not fit
Mac, 512 GBDoes not fit
DGX Spark 128 GBDoes not fit
Ryzen AI Max+ 395, 128 GBDoes not fit
CPU only, 64 GB DDR5-5600 dual channelDoes not fit

Same plan engine as the planner, with its default reserves and RAM. How plans are chosen.

Files exact Hugging Face bytes

ArtifactSizeFilesRevisionSource
Native safetensors mixed/custom native; ; stored tensor dtypes=F32/BF16/I64 · F
MultiverseComputingCAI/Comet_60B_A12B_BF16
114 GiB26aeb47274deHF files

Not found in the representative quant search (not proven unavailable): GGUF FP16, GGUF Q8_0, GGUF Q5_K_M, GGUF Q4_K_M, MLX 8-bit, MLX 4-bit, EXL3, AWQ, GPTQ. Complete selected file sets; projectors, MTP/draft and lookup files are listed separately and budgeted only by plans that use them. File bytes are not allocated memory.

Quality

No Artificial Analysis evaluation mapped to this checkpoint.

Arena text rating

Unknown

Publisher-reported benchmarks

None extracted from the model card.

AA and Arena describe the hosted model in the stated mode, not each quantized file; scores are never inherited by base, pruned, abliterated or quantized variants and never averaged together.

How to run it

vLLM patched nightly

source-documented, not executed source · 0.28.1rc1.dev628+g2a02f6efe plus bxyu/mtp-fix-try02 commit64edb1d9

export VLLM_USE_V2_MODEL_RUNNER=0
export VLLM_SSM_CONV_STATE_LAYOUT=DS

vllm serve MultiverseComputingCAI/Comet_60B_A12B_BF16 \
  --served-model-name model \
  --tensor-parallel-size 4 \
  --enable-expert-parallel \
  --trust-remote-code \
  --max-model-len 262144 \
  --gpu-memory-utilization 0.85 \
  --kv-cache-dtype fp8 \
  --mamba-ssm-cache-dtype float32 \
  --enable-prefix-caching \
  --no-disable-hybrid-kv-cache-manager \
  --block-size 128 \
  --mamba-cache-mode align \
  --enable-chunked-prefill \
  --max-num-seqs 128 \
  --max-num-batched-tokens 33920 \
  --async-scheduling \
  --enable-auto-tool-choice \
  --tool-call-parser qwen3_coder \
  --reasoning-parser nemotron_v3 \
  --speculative-config '{"method": "mtp", "num_speculative_tokens": 3}' \
  --limit-mm-per-prompt '{"image": 128, "video": 1}'

Recipes are source examples (settings are the publisher’s or author’s, not measured optima) and were not executed by LocalFit. A downloadable file does not prove your backend build supports the architecture.

Reported speeds 0

No reported measurement yet. Submit one.