Local AI Registry

Recipe

glm-5-2-int4-int8mix-dgx-spark-gb10-128gb-vllm-tp1

glm-5-2-int4-int8mix-dgx-spark-gb10-128gb-vllm-tp1
JSON APIFind compatible recipes

Observed LocalMaxxing leaderboard run. Evidence for compatibility, not an executable launch contract.

Record

Status
candidate
Source
localmaxxing
Engine
vllm
Accelerators
1
Tensor parallel
1
Context tokens
200,000
Max concurrency
1
chat
unknown
reasoning
unknown
tools
unknown
vision
unknown

Hugging Face model card

Identity

https://huggingface.co/QuantTrio/GLM-5.2-Int4-Int8Mix
Repository
QuantTrio/GLM-5.2-Int4-Int8Mix
Status
known
Link type
Exact Hub repository

Public Hugging Face repository confirmed by the Hub API.

Candidate: useful compatibility or speed evidence. The registry does not offer Run until promotion requirements are met.

Observed configuration

vllm

Evidence only · candidate · reference

Candidate evidence — not a Run contract

Source
https://www.localmaxxing.com/en/runs/cmr7eu1a7009tni01n3wqp3ti

Observed source tokens

Mechanical split of the source command. Unverified against the engine CLI. The registry does not offer Run for this recipe.

  1. vllm
  2. serve
  3. QuantTrio/GLM-5.2-Int4-Int8Mix
  4. --tensor-parallel-size
  5. 4
  6. --max-model-len
  7. 200000
  8. --max-num-seqs
  9. 6
  10. --speculative-config
  11. {method:mtp,num_speculative_tokens:4,draft_tensor_parallel_size:1,attention_backend:FLASHMLA_SPARSE}
  12. --tool-call-parser
  13. glm47
  14. --enable-auto-tool-choice
  15. --reasoning-parser
  16. glm45
  17. (4x
  18. DGX
  19. Spark
  20. GB10,
  21. RoCE,
  22. Bluey
  23. head
  24. :8210)
FlagValue
--tensor-parallel-size4
--max-model-len200000
--max-num-seqs6
--speculative-config{method:mtp,num_speculative_tokens:4,draft_tensor_parallel_size:1,attention_backend:FLASHMLA_SPARSE}
--tool-call-parserglm47
--reasoning-parserglm45

Measured speed

ConcurrencyContextPrefillDecodeTTFT msStatusSweep
1200,0002,99638.6873observedglm-5-2-int4-int8mix-dgx-spark-gb10-128gb-vllm-tp1-sweep

Remaining fields

Identity, launch, related records, and measured speed are shown above. This is the rest of the normalized record.

hardware count
1
hardware id
dgx-spark-gb10-128gb
id
glm-5-2-int4-int8mix-dgx-spark-gb10-128gb-vllm-tp1
model instance id
quanttrio-glm-5-2-int4-int8mix--int4-int8mix
recipe source
localmaxxing
schema version
local-ai-registry/v1
speed sweep ids
glm-5-2-int4-int8mix-dgx-spark-gb10-128gb-vllm-tp1-sweep
status
candidate

capabilities

draft launch

accelerator backend
nvidia
arguments
--model, QuantTrio/GLM-5.2-Int4-Int8Mix, --tensor-parallel-size, 1, --host, 0.0.0.0, --port, 8000, --max-model-len, 200000
container port
8,000
host port
8,000
image
vllm/vllm-openai@sha256:0a51ea5b4ae2dc5d81890e5173f54203d2a3ae0cfffe51b8fd2afd4391bfd967
ipc
host
kind
docker
shm size
16g
synthesized · generated at
2026-08-31T22:12:17Z
synthesized · image provenance
deepseek-fp8-rtx-pro-6000-blackwell-96gb-vllm-tp1
synthesized · template
vllm-openai-v1

mounts

read onlytarget
No/root/.cache/huggingface

engine

name
vllm

serving

max concurrency
1
max context tokens
200,000
tensor parallel
1
Provenance & metadata (3)

facts

capabilities.chat · provenance · captured at
2026-08-30T09:26:09Z
capabilities.chat · reason capability-not-verified
capabilities.chat · state unknown
capabilities.reasoning · provenance · captured at
2026-08-30T09:26:09Z
capabilities.reasoning · reason capability-not-verified
capabilities.reasoning · state unknown
capabilities.tools · provenance · captured at
2026-08-30T09:26:09Z
capabilities.tools · reason capability-not-verified
capabilities.tools · state unknown
capabilities.vision · provenance · captured at
2026-08-30T09:26:09Z
capabilities.vision · reason capability-not-verified
capabilities.vision · state unknown
engine.graph mode · provenance · captured at
2026-08-30T09:26:09Z
engine.graph mode · reason runtime-detail-not-published
engine.graph mode · state unknown
serving.kv cache tokens · provenance · captured at
2026-08-30T09:26:09Z
serving.kv cache tokens · reason kv-cache-capacity-not-published
serving.kv cache tokens · state unknown

metadata

localmaxxing · backend
cuda
localmaxxing · hardware label
GB10 Grace Blackwell
localmaxxing · notes
[STRUCTURED / JSON-output workload] GLM-5.2 UNPRUNED 744B (QuantTrio Int4-Int8Mix, all 256 experts) TP=4 across 4x NVIDIA DGX Spark GB10 over ConnectX-7 RoCE, MTP k=4 spec-decode, fp8 sparse-MLA (DSA) KV, full CUDA graphs, vLLM native multi-node (no Ray). Single-stream decode is MTP-content-dependent (temp 0, batch 1, warm): ~31 code / ~38 structured-JSON / ~24 prose. Aggregate ~60 tok/s at 6 concurrent. Recipe: https://github.com/tonyd2wild/GLM-5.2-QuantTrio-200K-4x-DGX-Spark
localmaxxing · observed command
vllm serve QuantTrio/GLM-5.2-Int4-Int8Mix --tensor-parallel-size 4 --max-model-len 200000 --max-num-seqs 6 --speculative-config {method:mtp,num_speculative_tokens:4,draft_tensor_parallel_size:1,attention_backend:FLASHMLA_SPARSE} --tool-call-parser glm47 --enable-auto-tool-choice --reasoning-parser glm45 (4x DGX Spark GB10, RoCE, Bluey head :8210)
localmaxxing · run id
cmr7eu1a7009tni01n3wqp3ti
localmaxxing · tokenized · arguments
vllm, serve, QuantTrio/GLM-5.2-Int4-Int8Mix, --tensor-parallel-size, 4, --max-model-len, 200000, --max-num-seqs, 6, --speculative-config, {method:mtp,num_speculative_tokens:4,draft_tensor_parallel_size:1,attention_backend:FLASHMLA_SPARSE}, --tool-call-parser, glm47, --enable-auto-tool-choice, --reasoning-parser, glm45, (4x, DGX, Spark, GB10,, RoCE,, Bluey, head, :8210)
localmaxxing · tokenized · fidelity
faithful

provenance

captured at
2026-08-30T09:26:09Z

sources

captured atkindurl
2026-08-30T09:26:09Znormalized-recipewww.localmaxxing.com/en/runs/cmr7eu1a7009tni01n3wqp3ti ↗