- Type
- Open weightsMIT License
- Params
- 754B
- Context
- 205K
active per word not recorded by us · about 154K words of context
Our take
Written Sep 17, 2026GLM 5.1 is a text model you can download under a licence that allows commercial use, changes and redistribution, but at 753.9 billion parameters it is not something you will run on your own machine. Most readers will reach it through a host, where it is best liked for coding and web-app prompts.
Use it for hosted coding and agent work where you want a model that is well liked on programming prompts and can pay per token. Its licence allows commercial use, changes and redistribution (MIT), so it is a reasonable base for a product. Skip it if you need to run the model on your own hardware, if your work is creative writing, or if you need measured speed or language coverage.
The case for it
- Coding and web-app building are its highest-rated categories on the Arena boards, both above its overall text score, so it is a sensible first trial for programming prompts.
- 23 hosted offers are listed, so you can reach it through an API without running it yourself.
- The licence allows commercial use, changes and redistribution (MIT), which keeps it usable in a product.
The case against it
- At 753.9 billion parameters with no smaller active-parameter figure, running it yourself needs hardware well beyond a single consumer machine.
- Recovery after a failed command is its weakest agent category, so agent work that needs to get back on track is the part to test first.
- Creative writing is its weakest measured text category, so prose is not where it shines.
How good is it?
An open-weight text model for everyday questions, drafting and coding, though it is less sure-footed with tools and picking up after a failed step.
- answering everyday questionsArena Text (overall) · 33rd of 168
- drafting and editing proseArena Creative Writing · 29th of 168
- writing and completing codeArena Coding · 34th of 168
- calling tools to carry out requestsArena Agent · Tool use · 49th of 55
- getting back on track after a step failsArena Agent · Recovery · 45th of 55
EverydayGeneral questions and everyday reasoning
Arena Text (overall)33rd of 168 · 1465
CodingWriting and fixing code on its own
Arena Coding34th of 168 · 1513
AgenticPlanning, calling tools, staying on task
Arena Agent35th of 55 · −0.022
Arena Agent is the only board that has scored it for this.
WritingDrafting and rewriting prose
Arena Creative Writing29th of 168 · 1448
Arena Creative Writing is the only board that has scored it for this.
Placings on Arena's agent boards, from live sessions people ran themselves. A model can lead on one of these and sit mid-field on the others.
Boards this model appears on that none of the ratings above are built on.
Every published score for this model12 scoresEvery figure we hold, from 12 boards, with who ran it and a link to the source — including the boards no rating above is built on.
Can you run it yourself?
GeForce RTX 4090 · 24 GB
Too large for this card. The weights do not fit even with part of them offloaded to system memory.
Apple M1 Pro (16-core GPU) · 32 GB
Too large for this card. The weights do not fit even with part of them offloaded to system memory.
Radeon RX 7900 XT · 20 GB
Too large for this card. The weights do not fit even with part of them offloaded to system memory.
Memory use by level
Against a 24 GB card.
What is quantisation? →This model on every device we track71 devicesThe Q4 build most people download, on each device: what the weights come to, how much context the memory leaves, and whether it runs. Smallest device that runs it first.
Check against your own machine → · Where to rent it hosted →
Or rent it from someone else
Prices checked between 4 hours and 7 days ago — each listing carries its own date.
- per 1M tokens
- $0.96 in / $3.03 out
- Context served
- 205K
- Throughput
- Not measured
| Provider | In / out per 1M tokens | Context | Throughput | Trains on prompts | Logs prompts | Zero retention |
|---|---|---|---|---|---|---|
| OpenRouterOpenRouter's own listing | $0.96 / $3.03checked 4 hours ago | 205K | not measured | Unknown | Unknown | Unknown |
| Baidufp8Through OpenRouter | $0.96 / $3.03checked 4 hours ago | 203K131K max reply | 44 tok/s | No | Yesunknown period | Unknown |
| StreamLakefp8Through OpenRouter | $0.97 / $3.04checked 4 hours ago | 200K128K max reply | 56 tok/s | No | Yesunknown period | Unknown |
| Chutesfp8Through OpenRouter | $0.98 / $3.08checked 4 hours ago | 203K66K max reply | 23 tok/s | No | Yesunknown period | Unknown |
| DeepInfrafp4Direct | $1.05 / $3.50checked 7 days ago | 203K | not measured | Unknown | Unknown | Unknown |
| SiliconFlowfp8Through OpenRouter | $1.19 / $3.74checked 4 hours ago | 205K131K max reply | 26 tok/s | No | No | Confirmed |
| AtlasCloudfp8Through OpenRouter | $1.26 / $3.96checked 4 hours ago | 203K182K max reply | 53 tok/s | No | Yesunknown period | Unknown |
| Alibaba Cloudfp8Through OpenRouter | $1.33 / $4.18checked 4 hours ago | 203K131K max reply | 53 tok/s | No | Yesunknown period | Unknown |
| PhalaThrough OpenRouter | $1.21 / $4.20checked 4 hours ago | 203K128K max reply | 23 tok/s | No | No | Confirmed |
| Novita AIfp8Direct and through OpenRouter | $1.38 / $4.40checked 4 hours ago | 205K131K max reply through OpenRouter | 33 tok/sthrough OpenRouter | DirectUnknownThrough OpenRouterNo | DirectUnknownThrough OpenRouterNo | DirectUnknownThrough OpenRouterConfirmed |
| Nebius AI Studiofp8Through OpenRouter | $1.40 / $4.40checked 4 hours ago | 203K182K max reply | 38 tok/s | No | No | Confirmed |
| FriendliThrough OpenRouter | $1.40 / $4.40checked 4 hours ago | 203K182K max reply | 43 tok/s | No | Yesunknown period | Unknown |
| Z.AIfp8Through OpenRouter | $1.40 / $4.40checked 4 hours ago | 203K131K max reply | 27 tok/s | No | No | Confirmed |
| GMICloudfp8Through OpenRouter | $1.40 / $4.40checked 4 hours ago | 203K182K max reply | 50 tok/s | No | Yesunknown period | Unknown |
| Venice AIfp8Through OpenRouter | $1.40 / $4.40checked 4 hours ago | 200K80K max reply | 34 tok/s | No | No | Confirmed |
Baidu
DeepInfra
Novita AI
Nebius AI StudioAcross the 15 listings we hold: 13 say they do not train on prompts (1 of them only through OpenRouter), 0 say they do and 2 do not say. 6 appear in the zero-retention registry we check (1 of them only through OpenRouter); the rest are unknown to us.
What each host's API supports
From the parameter list each endpoint publishes. Streaming is omitted: nothing we hold reports it, for any model.
| Provider | Tool calling | JSON output | Strict schema |
|---|---|---|---|
| OpenRouterOpenRouter's own listing | ✓ | ✓ | ✓ |
| Baidufp8Through OpenRouter | ✓ | ✓ | ✓ |
| StreamLakefp8Through OpenRouter | ✓ | ✓ | ✓ |
| Chutesfp8Through OpenRouter | ✓ | ✓ | ✓ |
| DeepInfrafp4Direct | |||
| SiliconFlowfp8Through OpenRouter | ✓ | ✗ | ✗ |
| AtlasCloudfp8Through OpenRouter | ✓ | ✓ | ✓ |
| Alibaba Cloudfp8Through OpenRouter | ✓ | ✓ | ✓ |
| PhalaThrough OpenRouter | ✓ | ✓ | ✓ |
| Novita AIfp8Direct and through OpenRouter | ✓ | ✗ | ✗ |
| Nebius AI Studiofp8Through OpenRouter | ✓ | ✓ | ✓ |
| FriendliThrough OpenRouter | ✓ | ✓ | ✓ |
| Z.AIfp8Through OpenRouter | ✓ | ✓ | ✗ |
| GMICloudfp8Through OpenRouter | ✗ | ✗ | ✗ |
| Venice AIfp8Through OpenRouter | ✓ | ✓ | ✓ |
Tool calling: 13 of 15 listings say yes, 1 says no, 1 publishes no parameter list. JSON output: 11 of 15 listings say yes, 3 say no, 1 publishes no parameter list. Strict schema: 10 of 15 listings say yes, 4 say no, 1 publishes no parameter list.
Models people weigh against GLM 5.1
When we formed this view
Recent changes
Each date is the day we first saw the change, or the day the maker announced it.
What we do not know about this model yet
- We hold no measured file for it, so all 3 sizes on this page are calculated from the parameter count.
- 1 of 15 listings publishes no parameter list, so what its API accepts is unknown to us.
- We do not hold the active parameter count for it, so how much of it runs on any one token is unknown to us.
- Nothing we hold says whether an endpoint streams, so we do not show it either way.
- 2 of 15 listings do not say whether they train on prompts, and 1 answers only through OpenRouter, not for its own listing.
- We hold no batch or off-peak rate for any of its listings.
- We hold a decode speed for it, but no prompt-processing (prefill) figure, so how long the input side of a job takes is unknown to us.
Licence and identifiers
What the licence allowsMIT License, what it allows commercially, and the identifiers you need to pull this model — its Hugging Face repo, our slug and a machine-readable card.
Licence
MIT License
Fully permissive: do anything with attribution. No patent grant, unlike Apache-2.0.
Identifiers
- Hugging Face
- zai-org/GLM-5.1
- Architecture
- Mixture of experts
- Takes in, gives back
- Text in, text out
- Catalogue slug
- z-ai-glm-5-1