Skip to content
Model Pareto

GLM-5.3 Flash NVFP4

NVIDIAOpen weightsNVFP4quantized from GLM-5.3 Flashhuggingface.co ↗
Compare with nearest rivalsCompare with…
Input
—
Output
—
Speed
~160tok/sest.
Context
1Mtokens

Rankings

Self-hostingEstimated

How we estimate →

What it would cost to run these open weights yourself on rented GPUs. No API sells this model, so this is its price on the chart.

Hardware
2× B200 180GB
NVFP4 weights
Throughput
~7,234 tok/s
many requests batched
Price
$0.18–1.62 /M tok
busy → light use
On your own machine
Mac Studio M5 Ultra 512GB (NVFP4, ~83 tok/s single-stream)
single consumer GPU or Mac
Assumptions (5)
  • NVFP4 weights (320B params, 18B active per token) + 40% KV-cache headroom ≈ 252 GB
  • 2× B200 180GB at $5.98–$14.24/GPU-hour on-demand (2026-09-24)
  • ~7,234 output tok/s aggregate at batch 207 (bandwidth-bound); MoE compute scales with active params
  • Blended 3:1 input:output; prefill ~118,125 tok/s
  • Low = 75% utilization at the low GPU price; high = 20% at the high price

Benchmark scores

~ italic, dashed = no published score yet, estimated from related benchmarks.

Coding

SWE-bench VerifiedEstimated
~#2 of 14 (estimated)~87.9%
  1. #1Claude Opus 596.0%
  2. ~#2GLM-5.3 Flash NVFP4Estimated87.9%
  3. #2Claude Sonnet 585.2%
No published score — estimated from related benchmarks
DeepSWE v1.1Estimated
~#15 of 32 (estimated)~66.1%
  1. #1GPT-6.1 Sol75.2%
  2. #2DeepSeek V4.1 Flash74.2%
  3. #14GPT-5.567.0%
  4. ~#15GLM-5.3 Flash NVFP4Estimated66.1%
  5. #15Gemini 3.8 Flash65.8%
No published score — estimated from related benchmarks
Terminal-Bench 4.0Estimated
~#16 of 85 (estimated)~26.2%
  1. #1Claude Sonnet 5.563.6%
  2. #2Claude Opus 5.559.6%
  3. #15DeepSeek V4.1 Flash26.8%
  4. ~#16GLM-5.3 Flash NVFP4Estimated26.2%
  5. #16Grok 4.725.8%
No published score — estimated from related benchmarks
LiveCodeBenchEstimated
~#6 of 25 (estimated)~88.7%
  1. #1Qwen3.8-Omni-Flash92.6%
  2. #2Qwen3.8 Flash-Next91.9%
  3. #5Nemotron 3 Ultra89.0%
  4. ~#6GLM-5.3 Flash NVFP4Estimated88.7%
  5. #6gpt-oss-120b87.8%
No published score — estimated from related benchmarks
AA Coding Agent IndexEstimated
~#7 of 15 (estimated)~54.4
  1. #1Claude Opus 5.566.0
  2. #2Claude Fable 5.162.2
  3. #6Grok 4.756.3
  4. ~#7GLM-5.3 Flash NVFP4Estimated54.4
  5. #7Muse Spark 1.354.3
No published score — estimated from related benchmarks

Writing & chat

MMMLUEstimated
~#3 of 5 (estimated)~86.7%
  1. #1Gemini 3.1 Pro92.6%
  2. #2Gemma 4 31B88.4%
  3. ~#3GLM-5.3 Flash NVFP4Estimated86.7%
  4. #3Claude Haiku 4.583.0%
No published score — estimated from related benchmarks
LMArena Text, Non-English (Elo)Estimated
~#23 of 54 (estimated)~1440 Elo
  1. #1Claude Opus 5.51501
  2. #2Claude Fable 5.1Few votes1496
  3. #22Hy31440
  4. ~#23GLM-5.3 Flash NVFP4Estimated1440
  5. #23GPT-5.4 mini1439
No published score — estimated from related benchmarks
LMArena Text (Elo)Estimated
~#23 of 54 (estimated)~1452 Elo
  1. #1Claude Opus 5.51509
  2. #2Claude Fable 5.11501
  3. #22Grok 4.61453
  4. ~#23GLM-5.3 Flash NVFP4Estimated1452
  5. #23Gemma 4 31B1451
No published score — estimated from related benchmarks
EQ-Bench Creative Writing v3 (Elo)Estimated
~#9 of 26 (estimated)~1966 Elo
  1. #1GPT-6 Astra2173
  2. #2Claude Fable 5.12162
  3. #8Grok 4.72007
  4. ~#9GLM-5.3 Flash NVFP4Estimated1966
  5. #9Muse Spark 1.31906
No published score — estimated from related benchmarks

Vision

MMMUEstimated
~#2 of 3 (estimated)~71.3%
  1. #1Claude Haiku 4.573.2%
  2. ~#2GLM-5.3 Flash NVFP4Estimated71.3%
  3. #2Llama 4 Maverick69.4%
No published score — estimated from related benchmarks
CharXiv ReasoningEstimated
~#5 of 9 (estimated)~81.2%
  1. #1Gemini 3.8 Flash86.2%
  2. #2Kimi K384.8%
  3. #4Qwen3.8-Omni-Flash83.5%
  4. ~#5GLM-5.3 Flash NVFP4Estimated81.2%
  5. #5Muse Glimmer78.8%
No published score — estimated from related benchmarks
OmniDocBench v1.5Estimated
~#4 of 8 (estimated)~87.6%
  1. #1MiniMax-M391.6%
  2. #2Kimi K391.1%
  3. #2Qwen3.8 27B91.1%
  4. ~#4GLM-5.3 Flash NVFP4Estimated87.6%
  5. #4Gemini 3.1 Pro85.3%
No published score — estimated from related benchmarks
GDP.pdf (AA)Estimated
~#17 of 43 (estimated)~16.9%
  1. #1GPT-6 Astra31.0%
  2. #2Muse Spark 1.326.6%
  3. #16Grok 4.617.2%
  4. ~#17GLM-5.3 Flash NVFP4Estimated16.9%
  5. #17Qwen3.8 27B16.6%
No published score — estimated from related benchmarks
ChartographyEstimated
~#10 of 23 (estimated)~28.6%
  1. #1GPT-6 Astra71.0%
  2. #2Claude Opus 5.566.3%
  3. #8Qwen3.8 Max (0902)29.1%
  4. ~#10GLM-5.3 Flash NVFP4Estimated28.6%
  5. #10Muse Spark 1.327.6%
No published score — estimated from related benchmarks

Hard reasoning

Humanity's Last ExamEstimated
~#44 of 182 (estimated)~31.1%
  1. #1Claude Opus 5.561.4%
  2. #2Claude Fable 5.159.1%
  3. #43Inkling31.9%
  4. ~#44GLM-5.3 Flash NVFP4Estimated31.1%
  5. #44A.X-K229.6%
No published score — estimated from related benchmarks
AIME (latest)Estimated
~#2 of 30 (estimated)~98.9%
  1. #1GLM-5.3 NVFP499.5%
  2. ~#2GLM-5.3 Flash NVFP4Estimated98.9%
  3. #2Inkling97.1%
No published score — estimated from related benchmarks