Up to date
108B parameters (8 of 128 experts per token), images and text in, text out, 128K tokens of context. 2 providers serve it from $0.30 per million input tokens.
Verified 2026-10-25 Rank #350 in LLMs · score 81.3 method v1 Sources 4 Downloads 30d 3.9k · #557 of 619 tracked
Picked by rule from 1 tracked piece: the lab's own post, the most shared, a field report, the latest.
GLM-4.6V accepts images and documents directly as tool parameters without converting them to text first.
Z.ai officially introduces the GLM-4.6V series, which integrates native Function Calling capabilities for the first time.
Live hosted endpoints, cheapest first, USD per million tokens. Ties break on output price, then provider name.
2 providers with a live price, from $0.30 to $0.30 per million input tokens; the median is $0.30. One dot per provider, cheapest in teal.
Hugging Face does not say who downloads a model or with what. These are the proxies: API providers listing it, quantized builds, fine-tunes/adapters/merges, and Spaces that load it.
Served by 2 API providers; 19 quantized builds for local use.
Top Spaces running it: akhaliq/anycoder ♥ 3,321 · hadadxyz/ai ♥ 819 · librarian-bots/dataset-card-drafter ♥ 2 · MohanamaniDeepika/anycoder ♥ 2
Weights only, an estimate: leave room for the KV cache. Method
About 65 GB at INT4, 129 GB at FP8, 259 GB at BF16: weights only, so leave room for the KV cache. Every expert must be resident, so total parameters count.
Estimate params × bytes per param × 1.2 · how we estimate
Weights, every dated article, and repo commits on one rail.
Teal is the weights (2025-12-07), amber triangles are 1 dated article (each links to it), grey ticks are repo commits.
25 derivatives on the Hub. Which build to run, by runtime: the most downloaded in each class.
From the Hugging Face model tree; quantized builds in teal, fine-tunes, merges and adapters in greys.
Everything else, folded: the full tables, every source.
Labels are evidence, not verdicts: measured a third-party leaderboard row for this model; author-reported a number the model author published in the card or eval files; structural a deterministic signal from the name, config or chat template, or a decision-model estimate. Each chip links to its source. All use cases.
Listed in LLMs
| Pipeline tag | image-text-to-text |
|---|---|
| Library | transformers |
| Architectures | Glm4vMoeForConditionalGeneration |
| Model type | glm4v_moe |
| Dtype | bfloat16 |
| Params by dtype | BF16: 107,710,927,360 · F32: 5,760 |
| Size band | XXL |
| MoE | yes |
| Experts | 128 |
| Experts per token | 8 |
| Context length | 131,072 (config.json max_position_embeddings) |
| Languages | zh, en |
| Papers | arXiv:2507.01006 |
| License | MIT (mit) |
| Hub snapshot | 2026-10-25 |
vLLM: supported · Glm4vMoeForConditionalGeneration · 7050ef881ff0835f6890a47db1859a24c65481d6SGLang: supported · Glm4vMoeForConditionalGeneration · 773cb938fe9484a8f7e7e46fb1b1fe96412c59ed
vLLM registry (7050ef881ff0835f6890a47db1859a24c65481d6) · SGLang registry (773cb938fe9484a8f7e7e46fb1b1fe96412c59ed) · Matched by architecture class name against each project's registry on main; 'supported' means the architecture is registered, not that this checkpoint was tested.
Weights-only memory for GLM-4.6V: params × bytes/param × 1.2 overhead. KV cache and activations are extra. How this is computed.
| BF16 (2 bytes/param) | 259 GB · Multi-GPU node (<640 GB) |
|---|---|
| FP8 (1 byte/param) | 129 GB · Multi-GPU node (<640 GB) |
| INT4 (0.5 bytes/param) | 65 GB · Single high-end GPU (<80 GB) |
Weights only (params × bytes/param × 1.2 overhead); KV cache and activations extra. MoE: all experts must be resident, so total parameters count.
quantized: 19finetune: 6
| Variant | Relation | Quant | Params | Band | License | Author | Likes | Downloads |
|---|---|---|---|---|---|---|---|---|
| cyankiwi/GLM-4.6V-AWQ-4bit | quantized | AWQ | 19.5B | L | mit | cyankiwi | 12 | 271.5k |
| unsloth/GLM-4.6V-GGUF | quantized | GGUF | — | — | mit | unsloth | 37 | 139.1k |
| ggml-org/GLM-4.6V-GGUF | quantized | GGUF | — | — | — | ggml-org | 7 | 71.2k |
| GadflyII/GLM-4.6V-NVFP4 | quantized | NVFP4 | 108B | XXL | other | GadflyII | 11 | 36.6k |
| mradermacher/GLM-4.6V-i1-GGUF | quantized | GGUF | — | — | mit | mradermacher | 1 | 30.3k |
| bartowski/zai-org_GLM-4.6V-GGUF | quantized | GGUF | — | — | mit | bartowski | 6 | 20.8k |
| mradermacher/GLM-4.6V-GGUF | quantized | GGUF | — | — | mit | mradermacher | 0 | 6.6k |
| noctrex/GLM-4.6V-MXFP4_MOE-GGUF | quantized | GGUF | — | — | — | noctrex | 2 | 3.1k |
| Beinsezii/GLM-4.6V-GGUF-HALO | quantized | GGUF | — | — | mit | Beinsezii | 5 | 2.7k |
| mlx-community/GLM-4.6V-mxfp4 | quantized | mxfp4 | 108B | XXL | mit | mlx-community | 4 | 2.4k |
| mlx-community/GLM-4.6V-6bit | quantized | — | — | — | mit | mlx-community | 0 | 1.6k |
| tiny-random/glm-4.6v | finetune | — | 2.79M | S | — | tiny-random | 1 | 1.6k |
| Beinsezii/GLM-4.6V-Q4F-Q8A-Q8SH-GGUF | quantized | GGUF | — | — | mit | Beinsezii | 1 | 1.3k |
| vivek51/GLM-4.6V-NVFP4 | quantized | NVFP4 | 61.5B | XL | other | vivek51 | 0 | 1.4k |
| cyankiwi/GLM-4.6V-AWQ-8bit | quantized | AWQ | 32.6B | L | mit | cyankiwi | 1 | 1.2k |
| unsloth/GLM-4.6V | finetune | — | 108B | XXL | mit | unsloth | 4 | 891 |
| unsloth/GLM-4.6V-FP8 | quantized | -FP8 | 108B | XXL | mit | unsloth | 1 | 805 |
| turboderp/GLM-4.6V-exl3 | quantized | exl3 | — | — | mit | turboderp | 8 | 316 |
| yujiepan/glm-4.6v-tiny-random | finetune | — | 2.79M | S | — | yujiepan | 0 | 667 |
| zeromodels/glm-4.6v | finetune | — | — | — | mit | zeromodels | 0 | 638 |
| jamiefutch/GLM-4.6V-MXFP4_MOE-GGUF | quantized | GGUF | — | — | — | jamiefutch | 2 | 442 |
| mlx-community/GLM-4.6V-nvfp4 | quantized | nvfp4 | 108B | XXL | mit | mlx-community | 0 | 526 |
| steampunque/GLM-4.6V-MP-GGUF | quantized | GGUF | — | — | mit | steampunque | 0 | 394 |
| blascotobasco/GLM-4.6V-96E | finetune | — | 82.8B | XL | mit | blascotobasco | 0 | 98 |
| blascotobasco/GLM-4.6V-112E | finetune | — | 95.3B | XL | mit | blascotobasco | 0 | 92 |
| Date | Commit | SHA |
|---|---|---|
| 2025-12-09 | Update README.md (#12) | 4e2d47e |
| 2025-12-08 | api link | 82f2a5c |
| 2025-12-08 | update | 387eda4 |
| 2025-12-08 | update | 833cb56 |
| 2025-12-08 | update | b8fe4c2 |
| 2025-12-08 | update | 1293c8b |
| 2025-12-07 | Update config.json | e5323a8 |
| 2025-12-07 | Add files using upload-large-folder tool | 0db1301 |
| 2025-12-07 | initial commit | 4ba8b9c |
Hosted endpoints observed for GLM-4.6V, sorted by input price (USD per million tokens). Each row names its observer and date; prices change daily, so check the source before quoting. Cheapest observed input price: $0.30/M.
| Provider | Model ref | Context | Quant | $/M in | $/M out | tok/s | TTFT | Tools | Status | Observed |
|---|---|---|---|---|---|---|---|---|---|---|
| Z.ai hf, openrouter | glm-4.6v | 128K | fp8 | $0.30 | $0.90 | 42.6 | 2883 ms | yes | live | 2026-10-25 |
| Novita openrouter | z-ai/glm-4.6v | 128K | bf16 | $0.30 | $0.90 | — | — | yes | live | 2026-10-24 |
Ranked by source signals (HN, Lobsters, authority domains, official posts) and recency. Summaries are generated from a stored copy of the article text; every sentence cites a span of that text, and nothing is shown that failed the citation check.
arXiv links found in the model card and its tags. "Peer-reviewed" is shown only with evidence: a published venue in OpenAlex (measured) or the authors' own journal reference on arXiv (reported).
Hub snapshot: 2026-10-25 · HF API: 2026-10-24 21:38 UTC · Providers: 2026-10-24 21:44 UTC · Articles: 2026-10-24 21:46 UTC · Summary: 2026-10-07 15:08 UTC · Last verified 2026-10-25
Email alerts are not open yet; the feeds above are the same data. See feeds for the item schema.
