Z.ai
GLM-4.5V carries the GLM-4.5-Air base into a vision-language model for text, images, and video. It supports up to 65,536 tokens of context, produces text, and combines multimodal reasoning with tool calls. The 106B total and 12B active parameters describe the MoE architecture.
GLM-4.6V is the newer multimodal iteration in Z.ai’s release line. GLM-4.5V remains a distinct earlier vision checkpoint with its own 65,536-token context and 8,192-token maximum output.
Profile
Published information about this model. Existing estimates are explicitly labeled.
Dated documentation and model-card observations. Provider limits, native context and extended context can differ. Configuration notes retain the source wording.
| Specification | Value and source |
|---|---|
| Weights | Open weightsSource |
| Access | API, LocalSource |
| Parameters | 106BSource |
| Active parameters | 12BSource |
| Context window | 65,536 tokensSource |
| License | MITSource |
| Architecture | vision-language MoE based on GLM-4.5-AirSource |
| Parameters | 106 BSource |
| Active parameters | 12 BSource |
| Rounded context label | 64kSource |
| Model configuration context | 65,536 tokens (Native checkpoint context; distinct from the rounded model-card label.)Source |
| Input | Text, Image, VideoSource |
| Output | TextSource |
| Reasoning mode | thinking, non-thinkingSource |
| Tool calling | YesSource |
| Release | 2025-08-11Source |
| Maximum output | 8,192 tokensSource |
These measurements have no matching peer values under the same test conditions. Their original values and sources remain available here.
Our model tests
Compare how the models respond to the same prompt. Each test shows the latest attempt, with no subsequent fixes to the generated code. Earlier runs remain available in the test details. These results do not contribute to an overall score.
A community classic for free-form SVG drawing.
GLM-4.5V
Reasoning (requested): high
Visually reviewed. This individual test does not establish overall model quality.
Generate an SVG of a pelican riding a bicycle
Token limit including reasoning: 8,192 tokens
Task origin (Simon Willison)Popularity
43.5 / 100
Model generations
Every relationship names its kind, its qualification and the dated source it rests on.
Evidence
Every statement links to its underlying documentation or leaderboard.
| Type | Evidence and data date |
|---|---|
| Speed | Median across serving providers, Non-reasoning. Output speed after the first streaming chunk, not total response time. Classified as deprecated by Artificial Analysis. Artificial Analysis maintains only the default 10K-input workload there. Retrieved September 8, 2026. |
| Price | Base API prices without caching or batch discounts. Higher context tiers and other rates are listed under Costs. |
| Input | Documented modalities. Retrieved September 8, 2026. |
| Output | Documented modalities. Retrieved September 8, 2026. |
| Research status | Research date September 8, 2026. 5 source URLs checked. This documents the inspected sources, not an exhaustive inventory of every publication. |
| Unresolved | detailed-benchmarks Original wording Card embeds a 42-benchmark image, but the referenced image returned 404 during capture; no scores were transcribed from an inaccessible image. |
| Unresolved | context-tokens Original wording Card states 64K for GLM-4.1V-9B and does not state GLM-4.5V context explicitly; 64K is not copied to this model. |
| Not found in the inspected sources | release-date Original wording Exact release date is published as 2025-08-11 in Project Updates; no later API snapshot date checked. |
| Additional source | Model metadata · GLM-4.5V model card · Editorial description reviewed October 2, 2026 |
| Additional source | Model metadata (retrieved September 8, 2026) · Z.ai GLM release notes · Editorial description reviewed October 2, 2026 |
| Additional source | Model metadata (retrieved September 8, 2026) |
| Additional source | OpenRouter: Popularity index (derived), CC BY 4.0 |
| Additional source | Vercel AI Gateway: Popularity index (derived), CC BY 4.0 |
| Additional source | Creative Commons Attribution 4.0 |