A detailed third-party breakdown confirms GLM-5.3-Flash launched on August 26, 2026, as a 320B mixture-of-experts activating 18B parameters per token, with MIT-licensed weights on Hugging Face from day one. It is the first natively multimodal model in the GLM-5 series and the first built on a new base since GLM-5.2, wi Pricing is set at $0.15 per million input tokens and $0.50 output, compared with $1.40 and $4.40 for the full GLM-5.3 flagship. Notably, GLM-5.3's own 744B open weights remain missing — Z.ai's Hugging Face organization has no GLM-5.3 repository — making GLM-5.3-Flash the only open-weight model developers can access tod

An NVIDIA DGX Spark community thread documents local inference testing of GLM-5.3-Flash across multiple quantization recipes. Users evaluated NVFP4-lab (mixed NVFP4/MXFP8-MTP), RedHatAI W4A4 compressed-tensors, and NVFP4-A16 variants, measuring decode tokens-per-second, time-to-first-token, and quality scores on tool-e The thread reveals a practical vLLM requirement: the framework needs a fix in modelopt.py because MTP layers are stored under language_model.model.layers.45, an issue already addressed in DeepSeek-V4-Flash support. Community members tested RedHat quantizations and discussed ongoing optimizations, with an "update coming

Yotta Labs' comparison piece documents GLM 5.3 Flash's August 26, 2026 release with MIT-licensed weights on Hugging Face, native image and video input, and Z.ai API pricing of $0.15 per million input tokens and $0.50 per million output tokens. The article reports a 1M-token advertised context window though evaluated at The piece highlights that the two launch tables barely overlap benchmark-wise: the flagship reported Terminal-Bench 3.0 (28.3, a six-fold jump over GLM 5.2's 4.6) while Flash reported Terminal-Bench 2.1 (84.3 versus Claude Opus 4.8's 85.0). On the shared DeepSWE v1.1 benchmark, Flash scored 63.4 against GLM 5.3's 66.9

A Medium technical analysis from Greek Ai describes GLM-5.3-Flash as Z.ai's newest open-weight release built around the principle of frontier-level capabilities without frontier-level inference costs. It confirms the 320B total / 18B active parameter architecture, native multimodal support, 1M-token context window, and The piece emphasizes that the GLM-5 family competition is shifting from raw parameter count to intelligence-per-compute, with GLM-5.3-Flash delivering dramatically lower inference costs. The model combines multimodal inputs, agentic capabilities, and a 1M-token context into a single architecture, making it suitable for

An Ampere comparison piece positions GLM-5.3-Flash against the full GLM-5.3 flagship. In independent Artificial Analysis testing, GLM-5.3 scores 60 on the Intelligence Index versus 57 for GLM-5.3-Flash, but Flash costs roughly one-ninth as much at normal API pricing and activates only 18B parameters versus 40B for GLM- Specs confirm GLM-5.3-Flash at 320B total / 18B active parameters with a 1M context window, while GLM-5.3 runs at 753B/40B active. GLM-5.3-Flash pricing is $0.15/M input and $0.50/M output versus $1.40/M and $4.40/M for GLM-5.3. GLM-5.3 is faster at generation (85 tok/s vs 49 tok/s) and wins on complex coding and long-

DataCamp's feature confirms GLM-5.3-Flash uses a 320B-A18B mixture-of-experts design with a 1M token context, scoring 84.3 on Terminal-Bench 2.1 — within striking distance of Claude Opus 4.8 (85.0) and behind GPT-5.6 Terra (87.4). The article reports a blended price of roughly $0.10 per million tokens, about a tenth of The piece notes GLM-5.3-Flash shipped first as the anonymous "Ox Alpha" preview before being publicly named, and flags practical deployment constraints: 49 tokens-per-second output speed and a 306 GiB FP8 checkpoint that rules out lightweight local hosting. DataCamp positions Qwen3.8-Flash-Next, which launched the same

The LLM Stats leaderboard ranks GLM-5.3-Flash at #17 overall with a composite score of 50.3 at a blended price of $0.17 per million tokens. It places in the "Good" tier for Tool Calling (7 of 193) and the "Average" tier for Reasoning (16 of 362), Chat (21 of 117), Vision (24 of 208), Coding (32 of 266), Math (38 of 327 Individual benchmark scores include GDPval-AA at 1773/3000 (rank 2, evaluated by Artificial Analysis), CharXiv-R at 0.89/1 (rank 8), and Terminal-Bench 2.1. The cost-efficiency chart positions GLM-5.3-Flash between DeepSeek-V4-Flash-0731 and GPT-6 Astra on a performance-versus-blended-price curve. Performance remains s

Local AI Zone's technical deep dive confirms GLM-5.3-Flash shipped on August 26, 2026, as a 320B-parameter MoE model activating only 18B per token, running natively in FP8 with a 1,048,576-token context window. It is the first natively multimodal model in the GLM-5 series, trained on a new base checkpoint rather than p The deep dive reveals that before launch, the model ran anonymously on OpenRouter as "Ox Alpha," topping usage charts and impressing observers including Patrick Collison. GLM-5.3-Flash beats GLM-5.2 across six coding and agentic benchmarks at roughly one-tenth the price and approaches Claude Opus 4.8 on long-horizon ag