Alex Getmansegnali precoci sull'IA
IT

GLM-5.3 Flash

Z.ai

Rilasciato Aggiornato
Finestra di contesto
1M
1.048.576 token
Output massimo
131K
131.072 token
Ingresso / Uscita
file, immagine, testo, video → testo

Prezzi API

USD / 1 milione di token

IngressoUscitaIngresso in cache
  • $0,15
zai-pricing
  • $0,50
zai-pricing
  • $0,03
zai-pricing
Calcola il tuo carico di lavoro

Cronologia

  1. · Versione dei pesi · link di riscontro
Prezzo ingresso/uscita · USD / 1 milione di token
ModelloLink di riscontroPrezzo ingresso/uscita
glm-5.3-flash
standard · testo
opencode-go$0,15 / $0,50
glm-5.3-flash
standard
zai-pricing
Percorso API non disponibile
$0,15 / $0,50
  1. · Link di riscontro · GitHub search · Z.ai · commits · zlxlabs/ci-infra-canary · @unknown
    Estratto
    Agent-Model: glm-5.3-flash Agent-Effort: high
  2. · Link di riscontro · GitHub search · Z.ai · commits · unixmonks/gpu · @unixmonks
  3. · Link di riscontro · GitHub search · Z.ai · commits · sgl-project/sglang-torchtpu · @Lumosis
  4. · Link di riscontro · GitHub search · Z.ai · commits · renj-arch/dimsred · @github-actions[bot]
  5. · Link di riscontro · GitHub search · Z.ai · commits · raidenyn/veles-android · @raidenyn-shadow
    Estratto
    the B5/v4 implementation and all integrations by actual ollama-cloud/glm-5.3-flash sessions (opencode-db safe metadata in package1-b5-v2-20261028/provenance/finish-phase-model-metadata.json); the earlier reviews were actually GPT (openai/gpt-5.6-sol high: windows-strict 520c2edd…,
  6. · Link di riscontro · GitHub search · Z.ai · commits · psyb0t/peen · @psyb0t
  7. · Link di riscontro · GitHub search · Z.ai · commits · phuchello2811-web/GLM · @doanbaphu
  8. · Link di riscontro · GitHub search · Z.ai · commits · opencrabs/opencrabs · @adolfousier
  9. · Link di riscontro · GitHub search · Z.ai · commits · nsbradley88/talus-public · @nsbradley88
  10. · Link di riscontro · GitHub search · Z.ai · commits · metaviews/jessehirsh.com · @metaviews
    Estratto
    z-ai/glm-5.3-flash as fallback, and a 404 'no provider serves this model'
  11. · Link di riscontro · GitHub search · Z.ai · commits · leshicorn/email-bot · @unknown
  12. · Link di riscontro · GitHub search · Z.ai · commits · juliolv/portfolio · @juliolv
    Estratto
    build (deepseek-v4.1-flash) -> audit (qwen3.8-flash) -> review (glm-5.3-flash as mandated) -> design (gpt-6-luna). The auditor is a different model family
  13. · Link di riscontro · GitHub search · Z.ai · commits · isotope13-dev/traits-dev · @unknown
    Estratto
    Model: opencode:zai-coding-plan/glm-5.3-flash@high
  14. · Link di riscontro · GitHub search · Z.ai · commits · isotope13-dev/traits-dev · @unknown
    Estratto
    Model: opencode:zai-coding-plan/glm-5.3-flash@high
  15. · Link di riscontro · GitHub search · Z.ai · commits · isotope13-dev/traits-dev · @unknown
    Estratto
    Model: opencode:zai-coding-plan/glm-5.3-flash@high
  16. · Link di riscontro · GitHub search · Z.ai · commits · ishan-parihar/artificial-route · @ishan-parihar
    Estratto
    breaker then cooled every healthy sibling model — glm-5.3-flash never failed yet was blocked by is_cooling_for ORing
  17. · Link di riscontro · GitHub search · Z.ai · commits · imoes/CentralStation · @unknown
    Estratto
    After the global LLM connector moved to llmbase.ai (z-ai/glm-5.3-flash), every tool call
  18. · Link di riscontro · GitHub search · Z.ai · commits · g00cey/tluna · @g00cey
  19. · Link di riscontro · GitHub search · Z.ai · commits · detrin/202611_AI_tinkers · @detrin
  20. · Link di riscontro · GitHub search · Z.ai · commits · databricks/unity-gateway · @lilly-luo
  21. · Link di riscontro · GitHub search · Z.ai · commits · brightmeows/bmsrs · @brightmeows
    Estratto
    Generated-By: pi 1.5.2 (glm-5.3-flash @ zhipuai-coding-plan, reasoning=max)
  22. · Link di riscontro · GitHub search · Z.ai · commits · brightmeows/bmsrs · @brightmeows
    Estratto
    Generated-By: pi 1.5.2 (glm-5.3-flash @ zhipuai-coding-plan, reasoning=max)
  23. · Link di riscontro · GitHub search · Z.ai · commits · brightmeows/bmsrs · @brightmeows
    Estratto
    Generated-By: pi 1.5.2 (glm-5.3-flash @ zhipuai-coding-plan, reasoning=max)
  24. · Link di riscontro · GitHub search · Z.ai · commits · brightmeows/bmsrs · @brightmeows
    Estratto
    Generated-By: pi 1.5.2 (glm-5.3-flash @ zhipuai-coding-plan, reasoning=max)
  25. · Link di riscontro · GitHub search · Z.ai · commits · agentgateway/agentgateway · @github-actions[bot]
    Estratto
    json.providers.nebius.models["zai-org/GLM-5.3-Flash"].reasoning.effort[2] = "max";
  26. · Link di riscontro · GitHub search · Z.ai · commits · KarpelesLab/oxinfer · @MagicalTux
    Estratto
    GLM-5.3: the text model of GLM-5.3-Flash (glm5_next) and GLM-5.3 (glm_moe_dsa),
  27. · Link di riscontro · GitHub search · Z.ai · commits · 183600/dyn_csa_experiment · @github-actions[bot]
    Estratto
    auto: iteration 1-bug [model: nvidia/z-ai/glm-5.3-flash]
  28. · Link di riscontro · GitHub search · Z.ai · pulls · vllm-project/vllm · @cjluo-nv
    Estratto
    by ModelOpt itself (NVIDIA/Model-Optimizer#3176), large MoE models (e.g. GLM-5.3-Flash), lm-eval accuracy, layerwise weight reload, EPLB, experts with biases.
  29. · Link di riscontro · GitHub search · Z.ai · pulls · vllm-project/vllm · @tuukkjs
    Estratto
    (open) edits the deprecated shared sizes. #63070 (draft, GLM-5.3-Flash) and #66320 (Qwen3.5) would move more models to breakable graphs on gfx950, which then get these
  30. · Link di riscontro · GitHub search · Z.ai · pulls · vllm-project/vllm · @jpvillam-amd
  31. · Link di riscontro · GitHub search · Z.ai · pulls · vllm-project/vllm · @kirillbilchenko
    Estratto
    compressed scale format (CSF) in GPU memory. On GLM-5.3-Flash with TP2, each rank uses 3.53 GiB less memory, and the KV cache holds 30.6% more
  32. · Link di riscontro · GitHub search · Z.ai · pulls · vllm-project/vllm · @cjackal
    Estratto
    In conjunction with #60229, support adaptive verification for GLM-5.3-Flash on Hopper GPUs by overriding `FlashInferMLASparseSM90Builder.get_varlen_cudagraph_max_query_len`.
  33. · Link di riscontro · GitHub search · Z.ai · pulls · vllm-project/vllm · @MIR-AMD
  34. · Link di riscontro · GitHub search · Z.ai · pulls · vllm-project/vllm · @HollowMan6
    Estratto
    [LoRA][Model] Add LoRA support for GLM-5.3-Flash <!-- markdownlint-disable -->
  35. · Link di riscontro · GitHub search · Z.ai · pulls · uwuclxdy/ai-pricelog · @cldbot
  36. · Post di Alex · famiglia · GitHub search · Z.ai · pulls · sgl-project/sglang · @ischencheng
    Estratto
    no longer fit next to the running requests. GLM-5.3-Flash NVFP4 on 2x B200 (TP2, the 0.9 split that was its default before #50018: 2,969,600 KV
  37. · Link di riscontro · GitHub search · Z.ai · pulls · sgl-project/sglang · @siliangchen-amd
    Estratto
    mHC models (the DeepSeek-V4 family, GLM-5.3-Flash) run the unfused mHC path: a chain of small kernels
  38. · Link di riscontro · GitHub search · Z.ai · pulls · sgl-project/sglang · @eppaneamd
  39. · Link di riscontro · GitHub search · Z.ai · pulls · sgl-project/sglang-torchtpu · @Bob-Chen222
  40. · Link di riscontro · GitHub search · Z.ai · pulls · sgl-project/sglang-torchtpu · @HarutMov
  41. · Link di riscontro · GitHub search · Z.ai · pulls · sgl-project/sglang-torchtpu · @lk-chen
    Estratto
    On GLM-5.3-Flash with attention DP8 (one TPU v7x 2x2x1 host, AgentX trace replay), round robin caches 0.22 of prompt tokens at c16, against a theoretical
  42. · Link di riscontro · GitHub search · Z.ai · pulls · sammasak/nixos-config · @sammasak
  43. · Link di riscontro · GitHub search · Z.ai · pulls · sadabaevb/ainek · @sadabaevb
    Estratto
    - **#84** is for `z-ai/glm-5.3-flash`: the same key and the same launcher our verifiers log, with release marks.
  44. · Link di riscontro · GitHub search · Z.ai · pulls · rodkisten/photointel · @renovate[bot]
  45. · Link di riscontro · GitHub search · Z.ai · pulls · renj-arch/dimsred · @renj-arch
    Estratto
    200 on 2026-11-13 (`nemotron-3-ultra`, `nemotron-3.5-super-vl`, `nemotron-3.5-lightning`, `nemotron-3.5-bolt`, `minimax-m3.1-flash`, `glm-5.3-flash`). Default is `nvidia/nemotron-3-ultra-550b-a55b`. Retired ids are recorded in the file as a comment with EOL dates,
  46. · Link di riscontro · GitHub search · Z.ai · pulls · pytorch/pytorch · @oonyshch
    Estratto
    | router softmax, 288 experts (GLM-5.3-Flash) | 16 tokens | 1.75x | 1.01x | 1.7x at 16384 tokens (14.1 to 8.1 us) |
  47. · Link di riscontro · GitHub search · Z.ai · pulls · misospace/courier · @itsmiso-ai
    Estratto
    separate reviewer session on the same model family (GLM-5.3-Flash, no model diversity available). Four findings fixed: ownership cross-validation lives in `charts/courier/templates/common.yaml.tpl` (not `secure.yaml.tpl`), isolation test
  48. · Link di riscontro · GitHub search · Z.ai · pulls · maximhq/bifrost · @sidbharth
    Estratto
    | `z-ai/glm-5.3-flash` | reasoning, fallback | NEAR AI-hosted reasoning model; too slow for the 30s chat-stream timeout |
  49. · Link di riscontro · GitHub search · Z.ai · pulls · lidge-jun/opencodex · @andrew05060414
    Estratto
    showed prices. `cn:deepseek-v4.1-flash` came from OpenRouter (0.3/1.2). `cn:kimi-k3.1-pro`, `cn:glm-5.3-flash`, `zd/gpt-6-sol` and `zd/gemini-4-argon` came from models.dev.
  50. · Link di riscontro · GitHub search · Z.ai · pulls · leptoskit/leptoskit · @ifiokjr
    Estratto
    Created on behalf of Ifiok Jr. (@ifiokjr), GLM-5.3-Flash (zai-coding-plan/glm-5.3-flash), high reasoning.
  51. · Link di riscontro · GitHub search · Z.ai · pulls · kollaborai/kollab · @kollaborai
    Estratto
    claude-code 71, codex-cli 46, antigravity 6, openai-oauth 67, glm-5.3-flash 12, glm-5.3 12.
  52. · Link di riscontro · GitHub search · Z.ai · pulls · kaina-visualsoftware/QA-Hub · @kaina-visualsoftware
    Estratto
    logs de produção mostram interrupções por `TimeoutError` no `z-ai/glm-5.3-flash`; houve também HTTP 502 no endpoint separado `qa-ai`.
  53. · Link di riscontro · GitHub search · Z.ai · pulls · harbor-framework/terminal-bench-science · @RishabhPatil
    Estratto
    (`harbor run`) with `gemini/gemini-3.1-pro-preview` via `gemini-cli` (5 trials), `zai-org/GLM-5.3-Flash` via `terminus-2` (5 trials), and `Qwen/Qwen3.8-27B` via `terminus-2` (5 trials), plus Oracle (`1.0` reward, 4/4 pytest
  54. · Link di riscontro · GitHub search · Z.ai · pulls · ggml-org/llama.cpp · @snakeice
    Estratto
    necessary step for running large MoE models (e.g. GLM-5.3-Flash 320B, 157 GB GGUF) on consumer hardware (32 GB RAM, single GPU) where CPU offloading makes
  55. · Link di riscontro · GitHub search · Z.ai · pulls · gcanlin/vllm · @gcanlin
    Estratto
    Vision tower alone (glm-5.3-bf16 weights, identical to GLM-5.3-Flash's BF16 tower), 1x B200, eager as served by default, FA4. Two interleaved runs per variant, mean of
  56. · Link di riscontro · GitHub search · Z.ai · pulls · figranium/fiptcha · @renovate[bot]
  57. · Link di riscontro · GitHub search · Z.ai · pulls · egonSchiele/agency-lang · @egonSchiele
    Estratto
    Anthropic stays at 20,000 from the provider layer; `z-ai/glm-5.3-flash` gets 131,072 from the model layer.
  58. · Link di riscontro · GitHub search · Z.ai · pulls · duketopceo/open-computer · @duketopceo
  59. · Link di riscontro · GitHub search · Z.ai · pulls · ddalcu/mlx-serve · @davidtai
    Estratto
    serve packs an outside quant plugin claims (DeepSeek-V4.1, GLM-5.3-Flash, Qwen3.8, Ling-3.1-flash)
  60. · Link di riscontro · GitHub search · Z.ai · pulls · crewAIInc/crewAI · @ats3v
  61. · Link di riscontro · GitHub search · Z.ai · pulls · cassandre60/ChessSRS · @cassandre60
    Estratto
    Model: z-ai/glm-5.3-flash
  62. · Link di riscontro · GitHub search · Z.ai · pulls · basetenlabs/baseten-cli · @sanieldalib
  63. · Link di riscontro · GitHub search · Z.ai · pulls · ashhart/TensorFold · @Bizuayeu
    Estratto
    On four GB10s at TP=4 (GLM-5.3-Flash XSFP4, `--context 1048576 --parallel 8`), round time in ms from our own grid client (`"draft": false`, greedy, 256 tokens,
  64. · Link di riscontro · GitHub search · Z.ai · pulls · ashhart/TensorFold · @Bizuayeu
    Estratto
    as they are. On four GB10s at TP=4 (GLM-5.3-Flash XSFP4, `--context 1048576 --parallel 8`), round time in ms from our own grid client (`"draft": false`,
  65. · Link di riscontro · GitHub search · Z.ai · pulls · ashcoft/document-mcp · @renovate[bot]
  66. · Link di riscontro · GitHub search · Z.ai · pulls · anomalyco/models.dev · @NachiketKandari
    Estratto
    v1 (2; new models should usually be v2_only): isoquant/glm-5.3-flash, isoquant/deepseek-v4.1-flash-1017-fast
  67. · Link di riscontro · GitHub search · Z.ai · pulls · anomalyco/models.dev · @models-dev-automation[bot]
  68. · Link di riscontro · GitHub search · Z.ai · pulls · anomalyco/models.dev · @jIrwinCline
  69. · Link di riscontro · GitHub search · Z.ai · pulls · ameetmadan/lautli · @renovate[bot]
  70. · Link di riscontro · GitHub search · Z.ai · pulls · alexsantossp71-lgtm/remedios_app · @alexsantossp71-lgtm
    Estratto
    `z-ai/glm-5.3-flash`, `stepfun-ai/step-5-preview`, `meta/muse-glimmer-30b`, `nvidia/nemotron-3.5-super-vl-120b-a12b` e `nvidia/nemotron-3.5-lightning-30b-a3b`. O app consulta o catálogo real da conta e só tenta os IDs retornados quando há correspondência.
  71. · Link di riscontro · GitHub search · Z.ai · pulls · alecpullen/marshal · @alecpullen
  72. · Link di riscontro · GitHub search · Z.ai · pulls · agentgateway/agentgateway · @github-actions[bot]
    Estratto
    json.providers.nebius.models["zai-org/GLM-5.3-Flash"].reasoning.effort[2] = "max";
  73. · Link di riscontro · GitHub search · Z.ai · pulls · TigreGotico/linguonnx · @renovate[bot]
  74. · Link di riscontro · GitHub search · Z.ai · pulls · SwarnimWalavalkar/noesis · @SwarnimWalavalkar
  75. · Link di riscontro · GitHub search · Z.ai · pulls · SwarnimWalavalkar/noesis · @SwarnimWalavalkar
    Estratto
    through the real TUI in a PTY with `opencode-go/glm-5.3-flash`. I captured the provider request bodies and read the SQLite rows instead of trusting the model's
  76. · Link di riscontro · GitHub search · Z.ai · pulls · SwarnimWalavalkar/noesis · @SwarnimWalavalkar
    Estratto
    - With the real model (glm-5.3-flash) in the TUI, one temporary home and two projects:
  77. · Link di riscontro · GitHub search · Z.ai · pulls · SwarnimWalavalkar/noesis · @SwarnimWalavalkar
    Estratto
    - With the real model (glm-5.3-flash) in the TUI, using temporary homes: a global Program ran by name from another project and recorded its exact
  78. · Link di riscontro · GitHub search · Z.ai · pulls · SwarnimWalavalkar/noesis · @SwarnimWalavalkar
  79. · Link di riscontro · GitHub search · Z.ai · pulls · SwarnimWalavalkar/noesis · @SwarnimWalavalkar
    Estratto
    - Real model over a PTY (`opencode-go/glm-5.3-flash`) in a temporary home, since deleted. In a first session I stated a codename and a rollback window.
  80. · Link di riscontro · GitHub search · Z.ai · pulls · ROCm/aiter · @eppaneamd
    Estratto
    - Callers that reach the new default: GLM-5.3-Flash at attention TP 2 (32 heads per rank), bf16 KV cache, gfx950,
  81. · Link di riscontro · GitHub search · Z.ai · pulls · ROCm/aiter · @eppaneamd
  82. · Link di riscontro · GitHub search · Z.ai · pulls · ROCm/aiter · @sumin-hong
  83. · Link di riscontro · GitHub search · Z.ai · pulls · ROCm/aiter · @sumin-hong
    Estratto
    every shape. This is what SGLang passes for GLM-5.3-Flash under CUDA graphs: the k-pool indexer's block-table capacity, 1M context / kpool 4 = 4,097 pages
  84. · Link di riscontro · GitHub search · Z.ai · pulls · QNFO/qnfo-workers · @rwnq8
    Estratto
    Draft for #6186: MODEL-AB-REGRESSION-1: qnfo-ai glm-5.3-flash -> standard on grounded-qa: win rate 0 < 0.4
  85. · Link di riscontro · GitHub search · Z.ai · pulls · Pb314314/vllm · @Pb314314
  86. · Link di riscontro · GitHub search · Z.ai · pulls · OpenHands/OpenHands · @sanchitmonga22
    Estratto
    (166 providers); `GET /api/llm/models?provider=wally` returns 10 models, including `wally/glm-5.3-flash` and `wally/deepseek-v4.1-flash`.
  87. · Link di riscontro · GitHub search · Z.ai · pulls · NousResearch/hermes-agent · @glitchbunny0
  88. · Link di riscontro · GitHub search · Z.ai · pulls · NVIDIA/hyperflare-inference · @aarongiuffre
  89. · Link di riscontro · GitHub search · Z.ai · pulls · NVIDIA/hyperflare-inference · @original-el8
  90. · Link di riscontro · GitHub search · Z.ai · pulls · NVIDIA/hyperflare-inference · @original-el8
  91. · Link di riscontro · GitHub search · Z.ai · pulls · NVIDIA/hyperflare-inference · @YouNeedCryDear
    Estratto
    The [GLM-5.3-Flash model card](https://huggingface.co/zai-org/GLM-5.3-Flash) documents `reasoning_effort` and `clear_thinking` and no way to turn thinking off. No revision of these three repositories' templates has contained `enable_thinking`.
  92. · Link di riscontro · GitHub search · Z.ai · pulls · NVIDIA/cudnn-frontend · @Anerudhan
    Estratto
    Add pooled DSA forward APIs for Blackwell GLM-5.3-Flash pools index keys before selecting attention tokens. Add forward APIs for gated cache updates, paged scoring, pool
  93. · Link di riscontro · GitHub search · Z.ai · pulls · NVIDIA/TensorRT-LLM · @ruocheng-nv
    Estratto
    end in #22082: a one-token prompt on the GLM-5.3-Flash FP8 target must not depend on the KDA slot's previous request, and the target is not
  94. · Link di riscontro · GitHub search · Z.ai · pulls · NVIDIA/TensorRT-LLM · @kaiyux
  95. · Link di riscontro · GitHub search · Z.ai · pulls · NVIDIA/Model-Optimizer · @shengliangxu
    Estratto
    - **GLM-5.3-Flash** (`models/glm5_next/modeling_ptq_vllm.py`, a new package): the k-pool indexer with its K-cache writers and DeepGEMM scorer, from `vllm_indexer.py`. Like `kimi_k3`, `glm5_next` has no `specs.py`: it
  96. · Link di riscontro · GitHub search · Z.ai · pulls · NVIDIA/Model-Optimizer · @cjluo-nv
  97. · Link di riscontro · GitHub search · Z.ai · pulls · NVIDIA-NeMo/Megatron-Bridge · @kamran-nvidia
  98. · Link di riscontro · GitHub search · Z.ai · pulls · Max-Health-Inc/config · @quotentiroler
    Estratto
    fix(vite): release notes default to Gemma 4 before glm-5.3-flash retires (4.5.1)
  99. · Link di riscontro · GitHub search · Z.ai · pulls · Max-Health-Inc/config · @quotentiroler
    Estratto
    - Changes the default model to `@cf/zai-org/glm-5.3-flash`, which followed the prompt rules more reliably than the previous default in testing. `RELEASE_NOTES_MODEL` still overrides it.
  100. · Link di riscontro · GitHub search · Z.ai · pulls · Lore-Hex/quill-router · @jasminelee
    Estratto
    Move the Fortanix GLM 5.3 Flash offer from the homepage to its model page and search text
  101. · Link di riscontro · GitHub search · Z.ai · pulls · Layr-Labs/flock-challenge · @yukon-autoresearch[bot]
    Estratto
    Model: zai-org/GLM-5.3-Flash Harness: Hermes Agent
  102. · Link di riscontro · GitHub search · Z.ai · pulls · Layr-Labs/d-inference · @jonathan308
  103. · Link di riscontro · GitHub search · Z.ai · pulls · Layr-Labs/arena-fortress · @yukon-autoresearch[bot]
    Estratto
    submitted game. The game model remains the challenge's `z-ai/glm-5.3-flash` through the organizer's gateway.
  104. · Link di riscontro · GitHub search · Z.ai · pulls · KarpelesLab/oxinfer · @MagicalTux
    Estratto
    - the text model of GLM-5.3-Flash (glm5_next) and GLM-5.3 (glm_moe_dsa),
  105. · Link di riscontro · GitHub search · Z.ai · pulls · IBM/Evoflux · @renovate[bot]
  106. · Link di riscontro · GitHub search · Z.ai · pulls · CYBERSZP0NT/szmelc-router · @sernik77
  107. · Link di riscontro · GitHub search · Z.ai · pulls · C2SM/icon4py · @egparedes
    Estratto
    models` and shows a red check. gt4py uses `zai-org/GLM-5.3-Flash` with fallbacks `Qwen/Qwen3.8-Flash-Next-FP8,Qwen/Qwen3.8-27B-FP8`.
  108. · Link di riscontro · GitHub search · Z.ai · pulls · BerryUIKI/CatCup-Tavern · @BerryUIKI
    Estratto
    ## Verified against the real endpoint (`deepseek-v4.1-flash` / `glm-5.3-flash`, 21 models)
  109. · Link di riscontro · GitHub search · Z.ai · pulls · BerriAI/litellm · @xinglan163
    Estratto
    feat(deepinfra): add GLM-5.3-Flash, MiMo-V2.6-Flash, Nemotron-3.5-Super-VL, and Hy4-preview
  110. · Link di riscontro · GitHub search · Z.ai · pulls · Augustas11/macprovider · @Augustas11
    Estratto
    M3 Ultra 256 GB, GLM-5.3-Flash, `msb-throughput --engine scheduler --prompt-tokens 1024 --decode-tokens 256 --runs 3`, the provider slice at `3675596d1` against main at `30ea73b89`, interleaved, two
  111. · Link di riscontro · GitHub search · Z.ai · commits · raspy135/tideLM · @unknown
  112. · Link di riscontro · GitHub search · Z.ai · commits · juliolv/portfolio · @juliolv
    Estratto
    paid model; the cheap tier (deepseek-v4.1-flash, glm-5.3-flash, qwen3.8-flash,
  113. · Link di riscontro · GitHub search · Z.ai · commits · islgit07/zcodebride-for-magpie · @islgit07
  114. · Link di riscontro · GitHub search · Z.ai · commits · 520042/aigw-gateway-panel · @520042

Segnali 1

Specifiche

GLM-5.3 Flash è un modello di Z.ai con pesi aperti, rilasciato il 2026-08-26. La finestra di contesto è di 1.048.576 token. L’output massimo è di 131.072 token. Il prezzo per milione di token è $0,15 per l’input e $0,50 per l’output.

Finestra di contesto
1.048.576
Output massimo
131.072
Ingresso
file · immagine · testo · video
Uscita
testo
Pesi
Aperti
Ingresso · USD / 1 milione di token
$0,15
Uscita · USD / 1 milione di token
$0,50
Ingresso in cache · USD / 1 milione di token
$0,03

Domande

Qual è la finestra di contesto di GLM-5.3 Flash?

La finestra di contesto è di 1.048.576 token. L’output massimo è di 131.072 token.

Quanto costa GLM-5.3 Flash?

Il prezzo per milione di token è $0,15 per l’input e $0,50 per l’output. L'ingresso in cache costa $0,03 per milione.

GLM-5.3 Flash ha pesi aperti?

Sì. Z.ai pubblica i pesi.

Quando è stato rilasciato GLM-5.3 Flash?

Il 2026-08-26.