GLM 5.3 Flash
Z.ai · Open weights · Released Aug 26, 2026 · z-ai/glm-5.3-flash
Best $0.04 in / $0.08 out via Relace
1.05M contextVisionToolsReasoningStructured outputPrompt cachingOpen weights
Available from 36 providersUSD / MTOK · PRICES 27 MINUTES AGO · SPEED 10 MINUTES AGO
EST. MONTHLY COSTclick a row for detail · click headers to sort
| PROVIDER | INPUT | OUTPUT | CACHE | CONTEXT | QUANT | TTFT | TPS | UPTIME |
|---|---|---|---|---|---|---|---|---|
 Relace | $0.04 | $0.50 | $0.013 | 1.05M | — | 0.93s° | 34° tok/s | 100.0% |
| OPOpenInference | $0.044 | $0.45 | $0.01 | 1.05M | fp4 | 2.6s° | 47° tok/s | 100.0% |
| SRSail Research | $0.045 | $0.60 | $0.029 | 1.05M | fp4 | 1s° | 111° tok/s | 99.4% |
| SRSail Research | $0.045 | $0.60 | $0.029 | 1.05M | fp4 | — | — | 99.3% |
 Pareto InferenceDIRECT | $0.06 | $0.20 | $0.012 | — | — | 2.4s | 63 tok/s | — |
 Reka | $0.06 | $1.60 | $0.04 | 262K | — | 0.94s° | 307° tok/s | 100.0% |
 StreamLake | $0.069 | $0.23 | $0.014 | 1.02M | fp8 | 2s° | 71° tok/s | 100.0% |
 InferenceNet | $0.07 | $0.08 | $0.034 | 1.05M | fp4 | ✕ | ✕ | 100.0% |
 DeepInfra | $0.075 | $0.25 | $0.015 | 1.05M | fp4 | ✕ | ✕ | 100.0% |
 CheaperInferenceDIRECTMARKETPLACE | $0.08 | $0.28 | — | — | — | 4.5s | 49 tok/s | — |
 Novita | $0.084 | $0.28 | $0.017 | 1.05M | fp8 | 4.2s° | 33° tok/s | 99.9% |
 GMICloud | $0.09 | $0.30 | $0.018 | 1.05M | fp8 | ✕ | ✕ | 99.9% |
 Wafer | $0.09 | $0.50 | $0.012 | 1.05M | — | 0.71s° | 63° tok/s | 100.0% |
 Decart | $0.093 | $0.31 | $0.019 | 1.05M | fp4 | 0.54s° | 59° tok/s | 100.0% |
| DEDekaLLM | $0.10 | $1.00 | $0.04 | 1.05M | — | — | — | 99.6% |
 Near AI | $0.105 | $0.35 | $0.025 | 1.05M | fp8 | ✕ | ✕ | 100.0% |
 Morph | $0.111 | $0.75 | $0.015 | 1.05M | fp8 | 1.4s° | 31° tok/s | 85.6% |
 Phala | $0.113 | $0.375 | $0.022 | 1.05M | nvfp4 | ✕ | ✕ | 100.0% |
 Inceptron | $0.12 | $0.55 | $0.099 | 1.05M | fp8 | ✕ | ✕ | 99.9% |
| CheaperInferenceDIRECTMARKETPLACE | $0.13 | $0.45 | — | — | — | 2.7s | 60 tok/s | — |
 Z.AI1ST PARTY | $0.15 | $0.50 | $0.03 | 1.05M | fp8 | 3.8s° | 43° tok/s | 100.0% |
 AtlasCloud | $0.15 | $0.50 | $0.03 | 1.05M | fp8 | 2.1s° | 45° tok/s | 69.2% |
 BaseTen | $0.15 | $0.50 | $0.03 | 1.05M | fp8 | ✕ | ✕ | 100.0% |
| BaseTen | $0.15 | $0.50 | $0.03 | 1.05M | fp8 | — | — | 99.8% |
 CoreWeave | $0.15 | $0.50 | $0.05 | 1.05M | nvfp4 | 1.1s° | 134° tok/s | 100.0% |
 Crusoe | $0.15 | $0.50 | $0.03 | 1.05M | fp4 | ✕ | ✕ | 99.7% |
 DigitalOcean | $0.15 | $0.50 | $0.03 | 1.05M | — | 0.94s° | 32° tok/s | 99.9% |
 Fireworks | $0.15 | $0.50 | $0.03 | 1.05M | — | 0.6s° | 158° tok/s | 99.5% |
 Friendli | $0.15 | $0.50 | $0.03 | 1.05M | — | 0.95s° | 146° tok/s | 100.0% |
 Modal | $0.15 | $0.50 | $0.03 | 1.05M | nvfp4 | ✕ | ✕ | 100.0% |
 Parasail | $0.15 | $0.50 | $0.03 | 1.05M | fp4 | ✕ | ✕ | 100.0% |
 SiliconFlow | $0.15 | $0.50 | $0.03 | 1.05M | fp8 | ✕ | ✕ | 73.4% |
 Together | $0.15 | $0.50 | $0.03 | 1.05M | — | ✕ | ✕ | 99.9% |
 Venice | $0.15 | $0.50 | $0.03 | 1.05M | — | 4.3s° | 20° tok/s | 99.1% |
| ParasailPRIORITY | $0.188 | $0.625 | $0.037 | 1.05M | fp4 | — | — | 99.9% |
| Fireworks | $0.225 | $0.75 | $0.045 | 1.05M | — | — | — | 99.7% |
TTFT/TPS: median (P50) over trailing 504h · chat10k profile (~10K in / 200 out) · measured against the provider API directly; ° = routed via OpenRouter (adds a hop) · ✕ = probed, endpoint returned no measurable stream · methodology v1.1 · measured 10 minutes ago
SERVE THIS MODEL AND NOT LISTED? LIST YOUR INFERENCE →
Input price history · best listed price, observed
↓ 20% SINCE AUG 31, 2026
$0.05 · AUG 31, 2026$0.04 · TODAY
Pricing variants · USD / MTok in / out
STANDARD
$0.04 / $0.50
per MTok in / out
BATCH
$0.06 / $0.20
async batch
CACHE READ $0.013
Recent changesRSS ↗
Oct 11, 2026price decreased · output0.2 → 0.08
Oct 10, 2026price increased · input0.032 → 0.04
Oct 9, 2026endpoint removed · cloudflare37 → 36
Oct 7, 2026endpoint removed · nextbit37 → 36
Oct 6, 2026endpoint removed · io-net37 → 37
Oct 5, 2026endpoint removed · primeintellect38 → 37
Oct 4, 2026endpoint added · primeintellect37 → 38
Oct 3, 2026price increased · input0.0263 → 0.035
About
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...
Specifications
Context window1,048,576
Max output943,717
Modalities intext, image, video
Modalities outtext
LicenseOpen weights
ReleasedAug 26, 2026
TokenizerOther
WeightsHugging Face ↗
Related
Compare with
Embed badge
[](https://modelindex.ai/models/z-ai/glm-5.3-flash)Sources
OpenRouter API27 minutes ago
AGGREGATORHugging Face27 minutes ago
COMMUNITYModelIndex benchmarks17 days ago
MEASUREDVerified 27 minutes ago · list prices, not negotiated rates.