Route order log in sign up
1 pick your primary route

Z.ai: GLM 5.3 Flash

z-ai/glm-5.3-flash · agent · text+image+video->text · 1M ctx · added 2026-08-26
$0.0423·0.3973 /M effective hit 84% 2.334T/d

GLM-5.3-Flash is a native multimodal model from Z.ai. It is suited for efficient coding and long-horizon agent tasks. Its hybrid sparse and linear attention architecture maintains accurate long-context behavior while...

view
openrouter/z-ai/glm-5.3-flash
Z.ai: GLM 5.3 Flash · agent · text+image+video->text · 1M ctx
2.334T/d 84% $0.0423·0.3973 – 2mo
sticker
life 429 tok/d dp age
¢ ¢ᵉ ✓
relace/fp4
$0.0399·0.0784·0.0300 $0.0305·0.0784ⁱ $0.0311·0.0776ᵉ 88% – – 2.5% ✓ 1494ms 36t/s 2mo ·
primeintellect
$0.0525·0.2500·0.0105 $0.0127·0.2500ⁱ $0.0196·0.2497ᵉ 78% – – 5.7% ✓ 1074ms 112t/s 1d ·
≈
morph/fp8
$0.0930·0.1220·0.0400 $0.0427·0.1220ⁱ $0.0468·0.1217ᵉ 79% – – 1.2% ✓ 742ms 52t/s 2mo ·
wafer/us
$0.0420·0.1840·0.0410 $0.0411·0.1840ⁱ $0.0413·0.1831ᵉ 66% – – 1.7% ✓ 505ms 48t/s 1d ·
⚡
open-inference/fp4
$0.0600·0.3000·0.0300 $0.0315·0.3000ⁱ $0.0330·0.2998ᵉ 90% – – 5.0% ✓ 728ms 154t/s 1mo ·
novita/fp8
$0.0840·0.2800·0.0168 $0.0203·0.2800ⁱ $0.0378·0.3369ᵉ 82% – – 1.7% ✓ 1556ms 52t/s 2mo ·
streamlake/fp8
$0.0885·0.2950·0.0177 $0.0214·0.2950ⁱ $0.0445·0.2942ᵉ 66% – – 0.2% ◐ 4681ms 38t/s 2mo ·
♻
inference-net/us
$0.1080·0.3500·0.0350 $0.0388·0.3500ⁱ $0.0388·0.3494ᵉ 95% – – 5.0% ✓ 437ms 84t/s 1d ·
gmicloud/fp8
$0.0750·0.2500·0.0150 $0.0181·0.2500ⁱ $0.0502·0.2502ᵉ 48% – – 0.2% ◐ 2751ms 30t/s 2mo ·
¢ⁱ
darkbloom/fp4
$0.0650·0.1800·0.0120 $0.0147·0.1800ⁱ $0.0583·0.1779ᵉ 13% – – <0.1% ◐ 2634ms 35t/s 1d ·
makora/fp4
$0.1400·0.2400·0.0300 $0.0357·0.2400ⁱ $0.0520·0.2390ᵉ 79% – – 1.4% ✓ 1034ms 70t/s 1d ·
coreweave/nvfp4
$0.1200·0.3800·0.0200 $0.0252·0.3800ⁱ $0.0375·0.3790ᵉ 81% – – 8.9% ✓ 698ms 94t/s 1mo ·
inference-net/fp4
$0.1080·0.3500·0.0350 $0.0388·0.3500ⁱ $0.0420·0.3488ᵉ 90% – – 2.8% ✓ 408ms 76t/s 1mo ·
reka
$0.1200·0.3000·0.0300 $0.0346·0.3000ⁱ $0.0490·0.2991ᵉ 78% – – 0.4% ✓ 722ms 71t/s 1d ·
wafer/nvfp4
$0.0650·0.2400·0.0640 $0.0641·0.2400ⁱ $0.0643·0.2379ᵉ 54% – – 0.4% ✓ 453ms 50t/s 1d ·
phala/nvfp4
$0.1125·0.3750·0.0225 $0.0271·0.3750ⁱ $0.0507·0.3747ᵉ 83% – – <0.1% ✓ 1160ms 55t/s ✕12h ·
tenstorrent/fp8
$0.0750·0.2500·0.0150 $0.0181·0.2500ⁱ $0.0650·0.2498ᵉ 26% – – <0.1% ✓ 5749ms 44t/s 1d ·
river
$0.1200·0.4000·0.0240 $0.0290·0.4000ⁱ $0.0489·0.3987ᵉ 76% – – <0.1% ✓ 709ms 75t/s ✕11h ·
★
together
$0.1500·0.5000·0.0300 $0.0362·0.5000ⁱ $0.0386·0.4995ᵉ 91% – – 22% ✓ 1044ms 114t/s 2mo ·
digitalocean
$0.1125·0.3750·0.0300 $0.0343·0.3750ⁱ $0.0529·0.3745ᵉ 74% – – 2.5% ✓ 1064ms 85t/s 2mo ·
near-ai/fp8
$0.0800·0.5000·0.0120 $0.0155·0.5000ⁱ $0.0391·0.4994ᵉ 65% – – 0.5% ✓ 755ms 22t/s 1mo ·
deepinfra/fp4
$0.1500·0.5000·0.0300 $0.0362·0.5000ⁱ $0.0409·0.4995ᵉ 90% – – 2.7% ✓ 1123ms 73t/s 2mo ·
modal/fp4
$0.1500·0.5000·0.0300 $0.0362·0.5000ⁱ $0.0409·0.4994ᵉ 89% – – 3.9% ✓ 435ms 75t/s 1d ·
mistral
$0.1500·0.5000·0.0300 $0.0362·0.5000ⁱ $0.0422·0.4996ᵉ 85% – – 2.2% ✓ 1580ms 77t/s 1d ·
particle-ai/nvfp4
$0.1200·0.5000·0.0300 $0.0346·0.5000ⁱ $0.0426·0.4994ᵉ 84% – – 5.1% ✓ 858ms 85t/s 1d ·
nextbit/fp4
$0.1200·0.4500·0.0300 $0.0346·0.4500ⁱ $0.0506·0.4490ᵉ 77% – – <0.1% ✓ 864ms 49t/s 1d ·
inceptron/fp8
$0.1750·0.5000·0.0300 $0.0375·0.5000ⁱ $0.0528·0.4585ᵉ 83% – – 0.9% ✓ 759ms 66t/s 1d ·
crusoe/fp4
$0.1500·0.5000·0.0300 $0.0362·0.5000ⁱ $0.0487·0.4995ᵉ 83% – – 2.4% ✓ 882ms 82t/s 2mo ·
baseten/fp8/us
$0.1500·0.5000·0.0300 $0.0362·0.5000ⁱ $0.0503·0.4995ᵉ 87% – – 0.2% ✓ 571ms 83t/s 1d ·
kt
$0.1500·0.5000·0.0300 $0.0362·0.5000ⁱ $0.0505·0.4996ᵉ 83% · – – 0.2% 5254.3M ◐ 1034ms 101t/s 11h ·
friendli
$0.1500·0.5000·0.0300 $0.0362·0.5000ⁱ $0.0509·0.4996ᵉ 78% – – 0.8% ◐ 479ms 124t/s 2mo ·
fireworks
$0.1500·0.5000·0.0300 $0.0362·0.5000ⁱ $0.0527·0.4994ᵉ 82% – – 7.4% ✓ 1047ms 88t/s 2mo ·
nebius/fp4
$0.1500·0.5000·0.0300 $0.0362·0.5000ⁱ $0.0550·0.4999ᵉ 73% – – <0.1% ✓ 1396ms 64t/s 1d ·
z-ai/fp8
$0.1500·0.5000·0.0300 $0.0362·0.5000ⁱ $0.0554·0.4997ᵉ 80% – – 5.8% ✓ 2684ms 34t/s 2mo ·
atlas-cloud/fp8
$0.1500·0.5000·0.0300 $0.0362·0.5000ⁱ $0.0561·0.4996ᵉ 80% – – <0.1% ◐ 2317ms 38t/s 1mo ·
venice
$0.1500·0.5000·0.0300 $0.0362·0.5000ⁱ $0.0595·0.4996ᵉ 73% – – 0.2% ✓ 655ms 69t/s 2mo ·
baseten/fp8
$0.1500·0.5000·0.0300 $0.0362·0.5000ⁱ $0.0601·0.4995ᵉ 76% – – 2.2% ✓ 730ms 81t/s 2mo ·
cohere/nvfp4
$0.1500·0.5000·0.0300 $0.0362·0.5000ⁱ $0.0651·0.4995ᵉ 67% – – 3.0% ✓ 1188ms 48t/s 1d ·
cloudflare
$0.1500·0.5000·0.0300 · $0.0362·0.5000ⁱ $0.0684·0.4989ᵉ · 68% · – – – · ◐ 1941ms · 9t/s · ✕1mo ·
sail-research/us
$0.1750·0.6250·0.0225 $0.0304·0.6250ⁱ $0.0559·0.6238ᵉ 51% – – <0.1% ✓ 4880ms 28t/s 1d ·
siliconflow/fp8
$0.1500·0.5000·0.0300 $0.0362·0.5000ⁱ $0.0713·0.4993ᵉ 67% – – 0.1% ✓ 1323ms 38t/s 2mo ·
sail-research/fp4
$0.1750·0.6250·0.0225 $0.0304·0.6250ⁱ $0.0581·0.6235ᵉ 64% – – <0.1% ✓ 6712ms 30t/s 1d ·
parasail/fp4
$0.1500·0.5000·0.0300 $0.0362·0.5000ⁱ $0.0766·0.4995ᵉ 58% – – 0.2% ✓ 2490ms 34t/s 1d ·
io-net/fp8
$0.1500·0.5000·0.0300 · $0.0362·0.5000ⁱ $0.0903·0.4996ᵉ · 50% · – – – · ✓ 1231ms · 25t/s · ✕1mo ·
parasail/fast
$0.1875·0.6250·0.0375 $0.0452·0.6250ⁱ $0.0985·0.6245ᵉ 48% – – 0.4% ✓ 841ms 123t/s 1d ·
modal/fp8
$0.4500·1.5000·0.0900 · $0.1086·1.5000ⁱ $0.2202·1.4982ᵉ · 64% · – – – · ✓ 546ms · 27t/s · ✕1mo ·
relace
$0.0900·0.3000·0.0180 · $0.0217·0.3000ⁱ – · – · – – – · · 1328ms · 38t/s · ✕1mo ·
wafer
$0.1000·0.3500·0.0200 · $0.0241·0.3500ⁱ – · – · – – – · · 685ms · 34t/s · ✕1mo ·
makora
$0.1400·0.4700·0.0240 · $0.0300·0.4700ⁱ – · – · – – – · · 733ms · 63t/s · ✕1mo ·
coreweave/fp8
$0.1500·0.5000·0.0500 · $0.0552·0.5000ⁱ – · – · – – – · · 690ms · 112t/s · ✕1mo ·
deepinfra/fp8
$0.1500·0.5000·0.0300 · $0.0362·0.5000ⁱ – · – · – – – · · 1336ms · 12t/s · ✕2mo ·
parasail/fp8
$0.1500·0.5000·0.0300 · $0.0362·0.5000ⁱ – · – · – – – · · 717ms · 15t/s · ✕1mo ·
phala/fp8
$0.1500·0.5000·0.0300 · $0.0362·0.5000ⁱ – · – · – – – · · 2520ms · 2t/s · ✕1mo ·
sail-research/fp8
$0.1500·0.5000·0.0300 · $0.0362·0.5000ⁱ – · – · – – – · · 1488ms · 10t/s · ✕1mo ·
nextbit/fp8
$0.2000·0.6750·0.0400 · $0.0483·0.6750ⁱ – · – · – – – · · 1326ms · 3t/s · ✕1mo ·
reka/fp8
$0.2250·0.7500·0.0450 · $0.0543·0.7500ⁱ – · – · – – – · · 1159ms · 14t/s · ✕1mo ·
vercel/zai/glm-5.3-flash
GLM 5.3 Flash · agent · text+image+video->text · 1M ctx
– 79% $0.0528·0.4500 6.4% 2mo
requesty/zai/glm-5.3-flash
glm-5.3-flash · agent · text+image->text · 1M ctx
– 90% $0.0668·0.5000 5.6% 2mo
hf/zai-org/GLM-5.3-Flash
GLM-5.3-Flash · agent · text+image->text · 1M ctx
– $0.1500·0.5000 – 2mo